How the rebels behind scholarly publishing app Mendeley—once labeled sellouts—are growing their company after being acquired by Elsevier.

In 2013, when Victor Henning announced that his six-year-old startup Mendeley would be acquired by one of the world's biggest media companies, he knew there would be blowback. He just couldn't have anticipated how bad it would get.

"Seeing that some of our most vocal advocates thought we had sold them out felt awful," Henning said recently over a tea in Amsterdam, where Elsevier, Mendeley's parent company, is headquartered.

Launched in 2007 by Henning and two friends at graduate school, Mendeley built an unlikely but very useful piece of software—think a variation on Evernote combined with Facebook—aimed at helping researchers organize their papers, annotate them, and share them with each other.

It swiftly took the academic world by storm. Researchers loved the ability to search for and in some cases access papers from journals they didn't subscribe to—a small protest against the billion-dollar industry that critics insist serves as a gatekeeper to the world's scientific findings. Within a few years, Mendeley had become an icon of the "open science" movement.

Enter Elsevier, part of the Dutch publishing conglomerate Reed Elsevier, it's one of the world's "big four" academic publishers. Amid challenges in other media sectors—and as academic library budgets shrivel up—these publishers have found a way to make handsome profits. In 2014, Elsevier listed a profit of $1.1 billion on revenues of 3.1 billion, largely by selling individual research papers and journal subscriptions.

Digital sales now make up two thirds of those profits, and by 2011, Elsevier had already begun mounting its own giant digital initiative. But it lagged behind in technology. And, like other large publishers, the company loathed Mendeley's open model; In 2013, it had forced Mendeley to remove its titles from its database. The thinking behind its acquisition of Mendeley—for a sum rumored to between $69 million and $100 million—was simple: to squash the threat Mendeley posed to its traditional subscription model, and to own the ecosystem that Mendeley had constructed, with its valuable data on the behavior of millions of researchers.

The purchase inspired revolt among its users. Mendeley was branded a sellout. Sean Takats, a professor of history and media at George Mason University and a director of Zotero, a Mendeley alternative that's popular among open science advocates, said the acquisition was akin to heresy. "I have nothing against making a buck," Takats told me. "But how can you partner with someone that's closed-science?"

When it came, the criticism stung Henning especially hard; even his former director of R&D, who had left the company shortly prior to the acquisition, joined in the chorus.

"I had steeled myself for some pretty violent reactions beforehand," Henning said. "After all, I was aware of Elsevier's reputation and the mistakes they had made."

Just a year earlier, thousands of researchers, led by Cambridge mathematician Timothy Gowers, boycotted Elsevier for lobbying Congress to keep research locked up behind paywalls for good, and for profiting from "the free labor of [researchers] and subscription fees from their institutions' libraries, for a service that has become largely unnecessary."

That service has helped Elsevier become one of the "big four" of academic publishers. Representing 3,057 journals, Elsevier is second only to recently merged Springer and Macmillan (which oversees the Nature publishing group) in terms of how many titles it controls.

For Elsevier, Mendeley's walled garden is a valuable new ecosystem, offering a lens into what researchers are working on and what papers they're looking for, and becoming a funnel, in essence, to sell more licenses to researchers, libraries, and academic institutions. In its latest annual report, published in March, the company highlighted "social collaboration through Mendeley" as it announced aims to use more data from Elsevier products, as well as third-party and customer data, to "expand content coverage" and "to increase content utility."

That arrangement, critics say, has led Mendeley away from the open science principles that it once symbolized. And now, full texts of paywalled research no longer circulate freely across its network, except within private groups.

"Once social networks start to get acquired or evolve, then you end up with a system where a large organization runs a 'walled garden,'" said Peter Murray-Rust, a lecturer at the University of Cambridge and an advisory board member at the Open Knowledge Foundation. "You can only get in or get information out by asking permission from the person who owns the garden. Mendeley was and is a walled garden."

Henning, however, insists he remains determined to meet the needs of a community that thrives on sharing ideas. He points to a revamped API that allows researchers to plumb its database, which is still available under a Creative Commons license. It now incorporates research from Elsevier's catalogs that Mendeley was once forced to remove. "We've kept the promises we made when we began," Henning contends.

"Our Relationship Was Schizophrenic"

The acquisition has taken the pressure off Mendeley to monetize, and helped keep the service free for individual researchers. It's also made possible long-awaited features, like improved recommendations and search, as well as an Android app.

And Mendeley keeps growing. Its user base—as of this week, at 4 million people worldwide—still uploads 1.6 million articles every day, with 100,000 new users joining every month, the company said. It counts Stanford, Harvard, and MIT as some of its big, institutional customers, and has doubled the size of its technical team, from 40 people to 85.

Still, even once it joined Elsevier, Henning admits, things weren't smooth sailing. He and the top of the Mendeley team worked hard to find allies within Elsevier to back his company's vision for open science. At 35, Henning is probably one of the youngest people working in the upper echelons of Elsevier, where he and Mendeley have had to prove they were more than just the "new kids." (In February, amid slightly lower year-on-year profits, the company announced it would "simplify its corporate structure.")

"Our relationship with Elsevier was schizophrenic," Henning said. "They had asked us to take their papers down from our network. But they sponsored several of our early events. They were supportive of us," he said.

In recent years, the publisher has been trying to retreat from its harsh image. The behemoth backed down on its support for proposed paywall-protecting legislation. And it has increased the number of open-access journals in its offering that people can read for free, but not any more significantly than other publishers' efforts. Henning also pointed to Elsevier's recent push for an industry-wide policy for publishers to allow sharing of copyrighted material within user groups on collaboration networks, as Mendeley does.

Long before big industries blamed the web for disrupting their business model—and before startups praised it for the same reason—scientists embraced it for its original purpose: to help researchers collaborate across borders and oceans, and create a repository of information that could be easily shared and searched. "Often," wrote Tim Berners-Lee, then an engineer at CERN, in his initial proposal for a World Wide Web, "the information has been recorded, it just cannot be found."

Nearly 20 years later, Henning and two classmates at the German Bauhaus-Universität Weimar, Jan Reichelt and Paul Föckler, noticed a similar problem: Research was largely siloed, and collecting it and sharing it was difficult. "I was looking at things from psychology, business, and philosophy," he said. "My advisor had to sit me down and tell me to stay focused on one thread of research and leave the other stuff as background knowledge."

In the beginning, Mendeley's biggest draw was that research groups could chat together in one place, share articles, and easily write papers together. It fostered collaboration in the cloud. In addition to allowing papers or their abstracts to be shared freely, Mendeley managed the entire workflow of researchers' existence more elegantly than anything had before. Researchers could access their profiles no matter which workstation they used, syncing their literature and notes to the cloud, much like Dropbox does now.

And Mendeley's open API—the first in the industry—freely offered metadata on the millions of papers that its users upload to the software, like titles, authors, keywords, and references. That allowed users to easily search through its database of papers, and allowed developers to do interesting things with that data. In 2010, this was a fresh, Web 2.0 idea in an industry that didn't have many of them.

Henning aggressively built a userbase by reaching out to open science advocates in the blogosphere and holding events where he preached a vision of making all types of research available to more people. Mendeley introduced new monetization strategies, like monthly fees for user groups and an institutional edition for entire universities, and raised funds to create the slick piece of software that it is today; investors included the backers behind Last.fm, Warner Music Group, and one of the founders of Skype. In just over four years, Mendeley raised about $12 million. In 2011, Mendeley had 800,000 users worldwide. By 2013, that number surpassed 2 million.

The grand hope was that big paper-sharing, social networking startups in the open science field—Mendeley but also sites like Academia.edu and ResearchGate—could eventually topple big science publishing stalwarts like Elsevier, much like the spread of MP3s had done to the music industry.

The publishing industry was not amused. In 2010, David Crotty, senior editor of Journals at Oxford University Press, marveled at Mendeley's "apparent naïveté towards copyright law." Keeping copyright-infringing papers off the service would be the right—and profitable—thing to do, he wrote, especially if Mendeley hoped to officially partner with universities. "It's not as sexy a path as being a Jolly Roger-hoisting rebel, but no one ever said that growing up was easy."

Open access advocates aren't pirates, they insist. Their aim, they say, is to accelerate the dissemination of important findings by removing the barriers that journals and their publishers put up.

"Open science is the process of designing your experiment in the open, recording everything openly," said Murray-Rust of the Open Knowledge Foundation. "Anyone with an interest should be able to take part. Ideally, anyone in the world could access a researcher's methods, materials, and findings. This can be done through an open social network, one that's not controlled by a commercial organization."

The bigger Mendeley got, the more scrutiny it got. Under the pressure of Elsevier and other big publishers, Mendeley had to scale back how openly research could course through its channels. Eventually the company removed Elsevier's abstracts from the API, as well as its PDF previews in its software.

Jason Hoyt, Mendeley's former head of R&D, didn't agree with the moves and left in 2013, explaining his move in a blog post. He would go on to start a family of open-access journals, called PeerJ, with the backing of venture capitalist Tim O'Reilly.

"There's a bit more to the story of why Jason left than he let on in his blog post. Either way, our relationship now is fine," Henning said. "We didn't have to 'justify' taking [Elsevier's data] down. After all, Elsevier owned the copyright, and we had to comply."

Three years later, the publishing giant came calling again, this time to buy. Henning was just 32. "It came to a point where such a large amount of our papers and data was connected to Elsevier," he said. "Our business was so closely tied with theirs."

Mendeley's investors were eager too. Sean Takats, who oversees Zotero, concedes that a corporate buyout was most likely Mendeley's best option to stay afloat. "The 'business model' could never continue on its own, let alone withstand the scrutiny of a public offering," he wrote at the time. For Mendeley, he wrote, "Acquisition was always the exit strategy."

In a blog post at the time in which he defended the acquisition, Henning wrote, "The opportunity to give our users access to better content, more data, and faster development was just too exciting to pass up." Elsevier "struck us as one of the most innovative and tech-savvy publishers out there." Olivier Dumon, managing director of academic and government markets for Elsevier, told The Times of London that "we're totally aligned when it comes to the product, the vision and the benefits this union will deliver to the research community."

After the acquisition was announced, William Gunn, Mendeley's head of academic outreach, summarized the reaction on his blog: "There are and will be a couple competing narratives: They bought us to bury us, we got paid tons of money so we said, 'Fuck Open Access,' etc. This is going to be put in the context of Google Reader shutting down, Delicious 'sunsetting,' etc." He wrote, however, that, "I'm not personally getting a pile of money from all this, and I never would have stayed unless I was convinced that they legitimately want to be part of the change to an open access publishing system."

While a number of users vowed to flee Mendeley for other reference management software, open source alternatives like Zotero remain scarce and underfunded. Scott Cowley, a PhD student social media and digital strategy at Arizona State University, described the problem in a comment on Danah Boyd's blog post: "Until there's a non-profit project that goes to bat for the academic community, I don't see being able to avoid tradeoffs whether in technology or philosophy."

Takats doesn't keep careful track of Zotero's user numbers, he said, so he can't be sure if there was a mass exodus from Mendeley to his software. Takats isn't focused on revenue: The project makes enough money to sustain itself by selling storage space, and he doesn't work on it full-time. It's just a side project, he said.

The Elsevier Era

Post-acquisition, Mendeley has relaunched its API, the stream of data that allows any developer to dig into its thickets of data; Henning is proud that Elsevier's PDF previews and abstracts are back in it. "Anyone is welcome to take our API and come up with a better system."

Still, open science advocates don't think an API makes Mendeley open enough. Said Takats: "There's a difference between open source and an open API."

To really reform research, Henning said he's taken up a new mission: Fixing the process of peer review, in which submitted articles are first carefully critiqued by researchers in the article's subject area before publication. Open access research hubs like arXiv and the Public Library of Science (PLoS) have sought to remedy this in recent years. In the near future, peer reviewers will be able to annotate papers and see the other reviewers' comments directly in Mendeley, before publication. "The editor of the journal would receive all of these comments by email and would just send them all to the paper's authors. The comments would overlap and sometimes contradict one another. We're eventually going to avoid this," Henning said.

Another volley against tradition could come in the form of a new Elsevier acquisition, a London-based news-tracking startup called Newsflo. Its software could offer Mendeley users a way to track their research's impact across the web, and become a new way to gauge success. This kind of alternative metric is one of open science's ways of rising up against the traditional journal citation system, where a journal's influence alone can make a researcher's career.

Another new initiative by Mendeley and Elsevier is also aimed at the open science community. This past December, Axon hosted a group of 15 science startups at the LeWeb conference in Paris. Elsevier covered the more than $1,000 conference fee for each startup that came.

Transcriptic, a Silicon Valley-based life sciences startup, attended LeWeb on Henning's invitation. Max Hodak, Transcriptic's CEO, notes that the life sciences business community is pretty small, and any opportunity to meet other startups is valuable. "I know these are people that will be doing important things down the line," Hodak said of the other participating startups.

One of the other startups that participated in the LeWeb conference through the Axon program was The Winnower, a Virginia-based science publishing startup. Josh Nicholson, its CEO and cofounder, doesn't even use Mendeley. He has spent most of his academic career using competitor EndNote.

But "it's a pain to use," Nicholson said. Soon, he thinks, he'll switch over to Mendeley.

After a series of missteps in video game development, this duo now hosts multi-player servers for the popular game Minecraft.

A two-person company from the small southern French town of Narbonne has unlocked a lucrative revenue stream from a global trend: Minecraft. After a series of missteps in video game development, Starlancer Studios, run by Jacques Vaquier and his business partner, Gregory Jung, now hosts multi-player servers for the immensely popular game.

"I'd spent a good amount of time playing video games. So, I was really motivated to develop video games myself," Vaquier says, using the French word impassioné, or impassioned, to express his enthusiasm for gaming. Initially, hosting Minecraft servers was only a way to fund his video game development business.

Taking advantage of Minecraft's worldwide success, Vaquier and Jung are expanding Starlancer Studios'Minecraft operations into North America and the rest of Europe this month. They will launch their new English-language site to promote their Minecraft servers on both sides of the Atlantic, on top of their three-year-old site for the French market.

Minecraft was created by Stockholm-based Mojang, which is now part of Microsoft. At the time I wrote this, Mojang reports there were 19,405,002 paying PC and Mac customers, and the number keeps growing. Known as a "sandbox" video game, its threadbare rules let players wander through virtual realms. The only real imperative is that a new player must gather enough resources to build a shelter by nightfall.

Minecraft servers enable multiple players to play the game together, without requiring people to keep their machines on at all times, or at least whenever their friends want to continue playing. The business opportunity lies in renting server space out to groups of players. And Minecraft isn't the only game that requires servers for more than one player to play against one another. Games like Call of Duty and Battlefield work under the same principle.

"We have 8,000 servers available to us in a data center in Paris," Vaquier says. "The value is that they are always online, are reliable, and are stable." He now has a new partnership with a data center in North America, to serve his new clients with the same dependability.

Minecraft's Global Takeover

In 2008, Vaquier started developing browser-based video games with a team in Montpellier, France. There, he met Jung, and with their joint investment, they transformed the group's project into a business. It eventually flopped.

Vaquier and Jung then struck out on their own, forming Starlancer Studios in 2012 and focusing on creating a more intricate browser video game. To raise funds for this effort, they hosted Minecraft servers. Vaquier says that was the same year in which the game "exploded." When the money started rolling in, the duo decided to make Minecraft the business's focus.

Given that Minecraft's player numbers keep booming, very little may hinder Starlancer Studios' growth—except for competition from other Minecraft servers. James Copeland, the founder of Canadian-based GGServers, says there are hundreds of thousands of companies offering server space for Minecraft players. Like Vaquier, Copeland is taking advantage of the game's worldwide user base by renting out servers in eight locations around the world, with 10,000 to 12,000 virtual servers online at any given time.

Currently, Vaquier says, France is the third largest market for multiplayer Minecraft, after the U.S. and Germany. So entering the U.S. was a natural move for Vaquier's growing company. But after an exhaustive market research study, he devised a careful initial strategy based on offering add-on applications for Minecraft and free server access.

"The U.S. and French markets are really different," says Vaquier. "France is advanced in that it's easier for players to find servers at no cost," adding that running servers in Europe costs less than in the U.S. He's decided to differentiate his business for the North American market by offering a "sizable amount" of free servers to players, giving them the opportunity to get used to other perks of the service.

According to Vaquier, there are two ways to make a Minecraft server business stand out:

Offer fun extras. In addition to server space, the team develops user interfaces and extra applications, called mods, to enhance the Minecraft-playing experience. Some popular mods let players build a complete Middle Earth or re-create Westeros from Game of Thrones. Vaquier and Jung have developed their own 3-D maps for various Minecraft worlds, as well as a VoIP application that allows players to communicate with other players in real time.

Make interfaces simple. "We make our user interfaces our number-one priority," says Vaquier. "We want our users to be able to easily sign up for a server, automating as many steps as possible." He mentions that many of his clients are children and teenagers, aged 8 to 16. They won't take the time to configure IP addresses on their own.

Now 32, Vaquier says he'll go back to developing his original video game concept, but there's too much going on in the Minecraft world to divert resources away from the hosting business. For now, serving the game's vast base of users is all the opportunity his company needs.

When grad student Erica Berman visited Paris in 1992, the Internet was barely a thing. Now she runs a growing online lifestyle brand.

Erica Berman didn't move to Paris to start a business. She initially landed there in 1992 in a way that probably sounds familiar: as a freshly minted journalism school graduate, curious about life's possibilities.

At the time, the Internet was barely a thing. But over the next two decades, Berman would leverage the Internet's growing capabilities to turn her Parisian lifestyle into what she says is a profit-making business. Today, Berman runs the online boutique vacation rental business Haven in Paris and cultivates the company's brand through its companion lifestyle blog, HiP Paris.

In the world of luxury travel, lifestyle is a commodity. Berman found her industry sweet spot by promoting her business as a lifestyle brand within a network of bloggers. Haven in Paris now has properties in Paris, Provence, Tuscany, and London, offering apartments and villas that range from $950 to $13,500 per week.

Now Berman has plenty to keep her busy, as she navigates tech's fast-changing landscape. When Berman launched Haven in Paris in 2006, she had much less competition than she does today. "There weren't responsive websites; there was no Airbnb," she says.

While Airbnb is a threat to her business, Berman maintains a competitive advantage by offering a personal-touch, high-quality service over what she sees as an impersonal, massive catalog of real estate. Each property on Haven is vetted, and a real live human always greets renters on location.

Still, the challenges can be daunting: For example, Google recently changed its search algorithm to favor responsive websites that specifically scale to mobile devices and tablets. Haven in Paris is now overhauling its site to better meet the demands of mobile devices and prevent Google from downgrading them in future search results.

Before starting her business, Berman briefly worked as a waitress in a Parisian wine bar, and then attended photography school. Eventually, she met a business partner and started a property rental business, which they ran together until 2004. Over those 10 years, Berman amassed the pool of contacts she needed to found Haven in Paris. Along the way, she's also learned some invaluable lessons for any entrepreneur—no matter where they start.

Websites Must Be Beautiful—And Responsive

At Haven in Paris's launch, Berman worked with an "amazing" designer who was also a friend. The partnership was beneficial in more ways than one. "I was really lucky in the sense that our developers did a really good job of maximizing the site for SEO," Berman says, adding that she rarely pays for SEO. Berman and her development team spent a lot of time choosing the right keywords to get search engines to work for them. And now that Berman is optimizing Haven in Paris's site for mobile, she sees an opportunity to refresh the company's image.

Learn The Language

After finishing journalism school, Berman spent a summer perfecting her French in one of Middlebury College's famed language immersion programs. "You're bound to have a more enriching experience communicating with the French in French," Berman says. "People appreciate it when you can communicate with them in their native language, even if they do speak English."

Build A Network, Even With Competitors

The Haven in Paris website has been a natural magnet for potential clients, but Berman also has a very extensive network in Paris: interior decorators, apartment hunters, real estate agents, bloggers, and friends. She loves to network and is open to building relationships with the right people, even if those are competitors. Berman and some of her competitors even send clients and properties back and forth to one another.

Don't Sell Your Product On Your Blog

"As soon as you try to make it sales-y, people are not interested," says Berman. "If you read the blog, we don't often talk about Haven in Paris," Berman says. The blog's small editorial team does, however, like to highlight certain neighborhoods where Haven in Paris has properties, and in those posts, the team will link to the available properties. But for the most part, the blog is its own entity.

In fact, HiP Paris has more traffic than Haven in Paris does. "People read the blog and get to know us," Berman says. "When they travel to Paris, they end up renting from Haven in Paris. Then when in Paris, they use the blog. It's a big circle." The blog tends to chronicle the places to see and things to do in Paris that would appeal more to a foodie than someone wanting to jet from one monument to another.

Berman has left her Parisian adventures behind, and is now based in Maine. She periodically travels to Paris to check in on all of Haven in Paris's properties, giving her personal attention to every decision that could affect her clients. More than offering her customers a comfortable place to stay, Berman hopes to offer them a gateway into a lifestyle.

You listening, California?

Last month Heineken tweeted that it had saved €75 million ($83.8 million) over the last six years by using less energy and water in its breweries. Like many people, I was intrigued. So I decided to find out how they did it and if others could follow their lead.

This wouldn't be possible for every company: The Amsterdam-based Heineken, after all, runs a huge operation. It has 160 breweries in 70 countries and is one of the top three beer producers in the world, alongside SABMiller and Anheuser-Busch. The amount of barley, hops, and water that Heineken's beer production requires not only commands a ton of energy but also impacts farming communities globally.

When I inquired about the savings figure at Heineken, a corporate relations representative there was adamant that I receive its 2014 sustainability report in the mail. That seemed a bit inefficient and not super green, but sure. I was surprised to find a small stack of recycled, Heineken-branded coasters, instead of a report intended for shareholders.

The first coaster read: "This is the sustainability report," and included instructions on how to use a mobile app called Blippar to scan either the coaster or a standard Heineken bottle to redirect the app to the company's online sustainability report. The message was clear: Heineken doesn't just want shareholders to know how green its business is; it wants to directly involve consumers in a conversation around sustainability.

It turns out that the €75 million savings come from Heineken's long-term sustainability program, "Brewing a Better World." Started in 2010, the program has been continuously monitoring the company's improvements in its sustainable business practices against 2008 benchmarks. Heineken will reassess its progress every three years until 2020, but 2015 is a major milestone year for the program.

Small Changes, Big Savings

By the end of last year, Heineken had already surpassed some of its "Brewing a Better World" program's 2015 and 2020 goals. Compared to 2008 levels, Heineken reduced its breweries' water consumption by 23%. Buoyed by this early success, Heineken has tightened its 2020 water consumption targets even more: from 3.9 liters of water per liter of beer to 3.5. In addition, Heineken reduced its breweries' CO2 emissions by 30% compared to 2008, beating its 2015 targets one year early.

Heineken is making some green investments, transitioning to using more solar- and wind-powered brewing equipment as well as using greener cooling agents in their breweries. According to Peter Jonkers, green brewery program manager at Heineken, operators in Heineken's breweries worldwide have been constantly measuring the energy consumption of their brewing equipment, trying to beat an internal "utility benchmark model."

Heineken's virtual model uses historical energy consumption data from all of Heineken sites, combined with local climate circumstances and existing brewery equipment conditions to simulate baseline energy usage levels. Heineken could then quantify the breweries' actual energy consumption intensities, versus the model's, into a cost-savings figure.

Though small changes are responsible for saving the €75 million. The biggest cost-saver, Jonkers says, is keeping on top of the equipments' maintenance. "The savings isn't a question of setting up new machines; it's really a question of good maintenance and engineering," says Jonkers. A lot of businesses looking for short-term profitability will take a shortcut by saving on maintenance costs, he says, but it causes bigger problems in the long run.

Along with reducing water consumption and reducing CO2 emissions, the "Brewing a Better World" campaign has set out to more sustainably source raw agricultural materials and advocate for consuming alcohol responsibly—the latter being most visible through its "Dance More, Drink Slow" campaign. This year, Heineken is widening the program's focus further into two new areas: cultivating its relationships with local communities and promoting health and safety wherever Heineken is present.

A Marketing Playland

Heineken knows its sustainability program's branding opportunity is much more valuable than any cost savings it will inspire. For perspective, saving on energy over the last six years amounts to little more than €1 million in savings per year for the company, which grossed €19 billion in revenue in 2014. My new coasters are evidence of Heineken's obsession with marketing.

This year, Heineken will receive the equivalent of the Palme d'Or for the advertising industry: the "Creative Marketer of the Year" award at the Cannes Lions International Festival of Creativity. Naturally, the company is milking the "Brewing a Better World" program for all of its marketing potential. Next up on Heineken's marketing horizon is its #Legendary7 campaign. (All of my Heineken coasters boast the #Legendary7 hashtag.)

The "Legendary 7" is Heineken's pick of seven farmers across Europe that Heineken buys its barley and hops from. Through its campaign, Heineken aims to engage consumers in the farmers' sustainable farming practices, mainly through a dedicated mobile app. There's even a selfie function.

"Sustainability is often seen to be complex and inaccessible for consumers," said Mark van Iterson, Heineken's global head of design, in a statement. "However, sustainability is at the heart of all that we do and we wanted to find a way to encourage consumers and all our stakeholders to easily engage with Heineken's Brewing a Better World programme."

The "Brewing a Better World" campaign's marketing strategy has recently drawn attention from the academic world. Forest Reinhardt and José Alvarez, professors at Harvard Business School, studied Heineken's "Brewing a Better World" initiative with HBS students this spring, based on a case study they wrote last year with researchers Tonia Junker and Daniela Beyersdorfer. In the study, they investigated how Heineken could better communicate how sustainably sourced its ingredients are, given that the supply chain isn't 100% sustainable. Heineken was flattered HBS studied its program, in any case.

Real Change

Moving forward, Heineken will focus more on implementing its sustainable sourcing strategy in the developing world, through a partnership with the UN Industrial Development Organization. Jonkers says Africa will be a focus. Currently, Heineken owns 56 plants in 20 countries there, and the African market makes up 20% of Heineken's business. The company's Growing Together campaign showcases some of the work it has already been doing in a few sites in Africa, boosting local farming practices.

"Brewing a Better World" may also be brewing a better Heineken. While Heineken has been reducing it water consumption and CO2 emissions since 2008, its production and revenue numbers have steadily increased in the same period. Last year, Heineken grossed €19.2 billion in sales and produced 18.1 billion liters (4.7 billion gallons) of beer. That's a 24% increase in beer volume since 2009, when the company took in €16.1 billion.

For the most part, my #Legendary7 coasters have served their main function, giving friends a place to put their drinks when they weren't holding them. But we actually did start talking about the environment, sustainability, and what it means to be "green." Chalk another one up for marketing gimmicks.

Plus, the devices will make money on their own.

At CES this past January, IBM researcher Veena Pureswaran described the company's joint plan with Samsung to get home appliances to exchange cryptocurrency with one another. The currency, called Ether, is similar to bitcoin, except that the traded commodity isn't directly related to a financial value. Instead, Ether's value is computing power.

What distinguishes the Ether and bitcoin cryptocurrencies from traditional money is the online system that records their every trade. Networks of people called miners use the software to collectively verify and record these cryptocurrencies' every trade. Like ever-growing strands of DNA, the currencies' digital addresses, called blockchains, store the details of each trade. Bitcoin and Ether run on their own software platforms, but in both cases, a blockchain makes the whole idea possible.

Letting connected devices barter computing power in the Ether cryptocurrency would address a basic issue of the coming Internet of Things era: paying for the cloud services that will allow devices to do useful things and talk to each other. Instead of relying on ads, fees from paying users, or selling data to third parties, hardware makers could use Ether as the basis for device-to-device communications and transactions.

Ether would blanket all of a connected home's devices in a network of code to minimize computing costs, reduce the scale of operation, maximize device longevity, and guarantee consumer privacy for all stakeholders in a connected future. In short, Ether could be the Internet of Things's wonder drug.

The Blockchain's Evangelist

The company that is helping IBM and Samsung bring connected objects into Ether is Ethereum. Founded in 2014 by a group of bitcoin enthusiasts—including a 2014 Thiel Fellow named Vitalik Buterin—Ethereum is promoting a more widespread use of blockchain technology: blockchain applications for everything. Using Ethereum's platform, coding a blockchain application should be as easy as creating a webpage in HTML.

Ethereum has built an open-source software platform that any coder can use to write blockchain-based applications. Additionally, anyone can join Ethereum's network and offer up a computer's computing power to the network. Then, the applications that are built on Ethereum's software platform run on the collective power of all the computers that have joined Ethereum's network, rather than in one centralized datacenter.

"Basically, what we're doing is building the world's biggest computer," says Stephan Tual, Ethereum's chief communications officer. He imagines a scenario where more and more people will volunteer their desktops' computing power to run programs that were built on Ethereum's software platform.

For Tual, the premise behind evangelizing the blockchain isn't to dissolve a centralized source of power for political reasons, like Bitcoin was conceived to do; it's to help businesses save the money and effort they would need to run these Internet of Things applications in the cloud. Ethereum is making the blockchain a reality for some of tech's biggest players.

"What Ethereum offers is really making it easy to build this stuff without needing to worry about cryptographic principles that are, quite frankly, obscure for most developers," says Tual.

A Really Slow, Really Big Computer

There are applications that don't lend themselves to blockchains, such as playing video games or running other applications that require similarly substantial computing resources. "Ethereum is a really slow computer. It's a very big computer that always tells the truth, which is quite interesting, and also has no downtime, which is another cool property it has. But it is pretty slow, compared to your laptop, for example. Because it's the lowest common denominator that determines how fast it's going to run," Tual says.

But blockchain technology is perfect for simple tasks that need to run constantly. Connected devices that perform crude, repetitive tasks, like monitoring temperature or turning lightbulbs on and off, would benefit the most from the blockchain's peer-to-peer networking.

IBM and Samsung have employed Ethereum in their joint Internet of Things project called ADEPT, which stands for Autonomous Decentralized Peer-to-Peer Telemetry. Along with Ethereum, the companies plan to use two other peer-to-peer applications in the project: Telehash for inter-device messaging and BitTorrent for inter-device file sharing. Ethereum will execute code between the devices.

"The reason that they need this stuff is that they don't want to pay," Tual says, laughing. "It's not altruism. They don't want to pay for its maintenance."

Ether's Value

Ethereum's network of computers is growing every day, as members are rewarded with CPU power for breathing life into Ethereum's web of applications.

And the Ether cryptocurrency holds clear financial value to the people who are developing the Ethereum platform. Last summer, the company made $18 million on a pre-sale of Ether. That money now pays the salaries of Ethereum's London-based management team, as well as its 30 developers in Berlin and Amsterdam.

While Ethereum will eventually spend that initial pre-sale fund, Tual hopes the company will reap the rewards of the foundation he says it's currently laying for the Internet of 20 years from now. He says that Ether will most likely always be associated with a monetary value as long as humans develop the code for the blockchain applications. So long as there is demand to run these applications, Ether's supply of available computing power will be valuable.

If ADEPT becomes reality, connected gadgets will also govern how much Ether their code will merit. "In doing so, devices can become revenue-generating opportunities in their own right," IBM's Pureswaran says.

The technology promises a new legal battleground for rights holders, printer developers, and hobbyist figurine makers.

It didn't take long for the email from Katy Perry's lawyers to arrive.

Just days earlier, during the Super Bowl halftime show, a disoriented shark-clad dancer on the left side of Katy Perry had become an instant Twitter sensation. As he watched the Left Shark meme take off, Fernando Sosa, a 3-D designer based in Orlando, saw an opportunity to do what he tends to do with popular and not-so-popular figures: turn them into figurines. His catalog of of-the-moment 3-D-printed people already includes dozens of lively characters, including Chris Christie, Vladimir Putin, and Florida governor Rick Scott as a chestbuster alien from the movie Alien. Within days of Perry's performance, he began to sell his Left Shark figurine on Shapeways, a central marketplace for 3-D-printed objects, where you can find everything from GoPro attachments for drones to 18-karat gold jewelry.

The following day, Shapeways heard from Perry's lawyers. They demanded that the site swiftly remove Left Shark from the site, claiming that Perry owned the copyright on the shark suit. Within hours, the site complied.

This wasn't the first time Sosa had struggled with intellectual property on the 3-D-printed frontier. In 2012, after he began selling a scaled-down replica of the iron throne from the HBO series Game of Thrones, which doubled as an iPhone dock, HBO sent him a cease-and-desist letter that asserted its rights to replicas inspired by the series. Sosa stopped selling the dock but quickly designed and started to sell a variation of it.

So far, the biggest lawsuits in 3-D printing have involved patent claims related to the design of the machines themselves. But the application of the technology could prove even more litigious. If the 3-D printers can allow hobbyists and small manufacturers to print anything—including an object that infringes upon a trademark or copyright, or a patented device—then they could be to the world of physical objects what the MP3 was to the record industry.

Now facing Perry's lawyers, Sosa wasn't deterred. "They sent a letter without any backing to pretty much scare small businesses out of the competition."

Familiar with the terrain, Sosa moved his Left Shark to Etsy and his personal website, then took to Twitter to drum up awareness of Perry's lawyers' accusations. As Sosa's tweets went viral, NYU law professor Christopher Sprigman, who specializes in intellectual property conflicts, offered to represent him.

Sprigman found two problems with the conduct of Katy Perry's lawyers: First, they didn't issue Shapeways a formal DMCA takedown notice. According to the Digital Millennium Copyright Act, it is the only tool that Shapeways is legally required to act on. Second, the copyright claim was erroneous to begin with. Perry didn't own the copyright on the shark suit, because the suit wasn't necessarily copyrightable to begin with. Costumes fall into the category of "functional goods," legal jargon for something that serves a useful purpose, as opposed to just being ornamental, Springman explains. Within two weeks of Perry's initial threat, Sosa's Left Shark was back up on the site.

It isn't clear yet if the rapidly growing technology of 3-D printing is straining other potentially competitive industries. At the moment, the NPD Group reports that action figures, accessories, and role-play toys accounted for $1.26 billion of $20 billion in U.S. toy sales in 2014. From now until 2018, the market analysis firm Gartner predicts the global $84.1 billion toy industry will grow by 4.2%. Over the same period, the global 3-D printing market is projected to reach $13.4 billion in sales, achieving a 103.1% compound annual growth rate.

Of course, a 3-D printer isn't exactly like an MP3. Unlike a digital copy, a three-dimensional copy possesses a permanence that outstrips a digital file. Michael Weinberg, Shapeways's general counsel and former vice president at Public Knowledge, a digital advocacy group, says that not every 3-D printer file is necessarily protected by copyright. "Some will be, but others will be protected by patent instead. Patent protection is different from copyright protection in a number of ways, and we won't simply be able to assume that the rules of thumb that governed what can be done with an MP3 will apply cleanly to 3-D printed stuff." This is uncharted territory.

Amid Accusations, Companies Fight, Eventually Adapt

The legal risks of 3-D printing have sent companies looking for solutions. Last summer, for instance, Shapeways introduced a revenue-sharing scheme with the toymaker Hasbro, to allow its makers to legally sell My Little Pony figurines. Printer companies, meanwhile, might opt for technological approaches, like the digital rights management, or DRM, that's used to prevent unauthorized copying of software, music, and films. The law—specifically US Code 1201—makes it a crime to thwart so-called "technological protection measures," or TPMs, that are embedded in certain devices like 3-D printers. In November, Weinberg and his colleagues at Public Knowledge filed a request for an exemption to the law, with the aim of decriminalizing the "jailbreaking" of certain 3-D printers, so that tinkerers can legally experiment with materials not typically approved by the manufacturer.

In March, 3-D printer maker Stratasys, which owns Makerbot, tried to block the request on the grounds that this could allow people to "print . . . work that infringes on copyright." Public Knowledge claimed that prohibiting users from printing objects with certain materials and not others would disrupt innovation. In its opposition, Stratasys's lawyers claimed that the petitioners hadn't proved that innovation would be curtailed. After all, they argued, hobbyists can still build their own open-source 3-D printers capable of printing anything. Stratasys has taken similar steps before: in 2012, it revoked the lease on Cody Wilson's printer after the gunsmith began printing parts for a pistol. The company did not respond to Fast Company's request for comment.

The Unlocking Technology Act of 2015, introduced by California Congresswoman Zoe Lofgren, aims to partly resolve this tension. The first version, passed and signed into law in 2013, let people unlock their phones when carriers wouldn't. The new act would add a clause to U.S. Code 1201 that would only ban tampering that is being done "in order to infringe or facilitate infringement of a copyright in a work protected under this title." The House Judiciary Committee is currently considering the act, after which it would make its way to the rest of Congress for a vote.

This argument—between ownership and innovation—is a familiar one in digital copyright battles. When Napster lost its case against the music industry in 2001, a consortium of law professors sided with Napster, writing: "The district court's ruling would ban a new technology in order to protect existing business models, and would invoke copyright to stifle innovation, not to promote it." In any case, the music industry only temporarily succeeded in quelling online music sharing around the world. Later, after Viacom asked Google for $1 billion in damages, it ended its copyright lawsuit with YouTube without a cash settlement. Meanwhile, new file sharing sites, the iTunes Store, and streaming technology would come to be the music industry's best hope, and it capitulated to a format it had once tried to destroy.

Even 3-D printing objects at home, for personal use and not for sale, could be infringing on copyright law. As a general rule there isn't a home or personal use exception to copyright or patent law, Weinberg explained. "In other words, just because you make a copy privately does not release you from liability if you are infringing. And in many cases making an unauthorized copy of an object will be infringing."

The "Success Kid" Meme

This narrative may contain an example for 3-D designers as they continue to create their wares in spite of mounting legal anxiety. Rather than fight over ideas, they are increasingly partnering with the Katy Perrys of the world, who might otherwise complain about them. And they are also collaborating with much smaller artists who could benefit from the exposure and the profits.

Ryan Kittleson, a 3-D designer who sells printed versions of his work on Shapeways, found an unlikely hit in an Internet meme called Success Kid. It's based on a photo in which a toddler stares directly at the photographer while crushing a fistful of sand in his clenched hand. Kittleson has sold several thousand Success Kids since he began selling them on Shapeways in the fall of 2012.

But as Success Kid took off in 2013, an agent contacted Kittleson to take it down. The agent, it turned out, represented the toddler's mother, who took the original photo that kickstarted the meme. The mother owned the photo's copyright, and Kittleson had infringed on it, the agent claimed.

"I didn't stop to consider who took the picture of Success Kid," says Kittleson. In the end, Kittleson struck up a deal to share the profits on his figurine with Success Kid's mother. She agreed, and her agent stopped hassling him.

Yet Kittleson's 3-D printed memes keep bringing up questions. Last year, Kittleson wanted to create a figurine of a twerking Miley Cyrus, but Shapeways refused to let him sell it. "I believe that it was due to Miley being the owner of her name and likeness, so I couldn't just sell a figurine of her," Kittleson says. Instead of doing away with the design completely, Kittleson replaced Cyrus's head with Albert Einstein's. He says he did it to show how ridiculous it was that he couldn't just sell the original twerking Cyrus.

Now, Kittleson approaches his new designs with more caution. If he uses a small artist's work for inspiration, he will approach the artist to strike up a revenue-sharing agreement first. If the artist doesn't agree, then Kittleson won't move forward with his 3-D rendering. But when it comes to dealing with creative sources that are backed by larger entities, he simply hedges his bets.

"For larger IP-owning companies, where there's no chance of them talking to a little guy like me, I will sometimes make an item anyway and take the chance that they might send me a DMCA takedown notice, and just hope that I fly under their radar," Kittleson says.

One could argue that Sosa's Left Shark and Kittleson's Success Kid were examples of fair use: fan art or parody. But proving fair use is not straightforward. As a case in point, the Associated Press sued Shepherd Fairey for using an AP photographer's 2006 portrait of then Senator Barack Obama for his famous 2008 "Hope" poster. The street artist claimed fair use, but ended up privately settling with the AP photographer. It seems that the easiest path to take is settling with the content creator directly, rather than redefining fair use whenever an intellectual property conflict erupts.

Fernando Sosa says he tried to work out a profit-sharing agreement with Katy Perry's camp, but to no avail. And in a bizarre twist, Perry's lawyers filed a trademark application for the Left Shark design and name two days after sending Shapeways the cease-and-desist letter. (They also sought protection for "Right Shark,""Drunk Shark," and "Basking Shark.") Their application even used an image of one of Sosa's Left Shark 3-D renderings, in what Sosa calls a blatant infringement on his copyrights. In April, the U.S. Trademark Office issued an initial rejection of Perry's application for protection of the Left Shark design, though it appeared to accept her attempt to register the word mark "Left Shark."

Sosa is prepared to field more attacks from Perry's legal machine, as its copyright claim is still up in the air: Perry's lawyers haven't yet responded to Sprigman's most recent opinion letter. Sosa says Perry has failed to prove copyright after four months.

While Sosa doesn't stand to make a lot of money on Left Shark—he has only sold around 200 of his figurines—he wants to defend his right to sell his design. He now has a Florida-based attorney on retainer, in case Perry's attorneys decide to strike back. To help defend against any future litigation, Sosa is raising funds through a crowdfunding site. So far, he has raised over $1,000 of his $2,000 goal.

"I believe that our whole copyright system needs to be overhauled so small entrepreneurs can enforce their copyright claims," Sosa says. "Not just wealthy singers with multimillion-dollar law firms."

The Growing Role of Maker Marketplaces

Like Stratasys, the printer maker, the marketplaces that host 3-D designs are playing a growing role in ensuring their users comply with the law, largely in order to protect themselves from giant liabilities and legal fees. YouTube famously relies on its terms of service to cite the Digital Millennium Copyright Act's safe harbor clause when a company accuses it of knowingly contributing to piracy.

Etsy, whose marketplace is central to the overall maker community, spells out intellectual property guidelines in its terms of service, with which the company expects users to comply once they open up a shop on the site. "We take intellectual property and copyright concerns very seriously, and we comply with the DMCA and remove items when we have proper notice," a spokesperson for Etsy wrote to Fast Company in an email.

Yet expecting every seller to truly understand the terms, and much less read them, is implausible, so community engagement is key. YouTube uses a video where hand puppets discover the world of copyright law.

Shapeways has taken community engagement in the 3-D design community a step further by creating a new business model. Where the rise of Internet communities once gave fan art a digital home, the maker community is spilling fan art into the physical world.

Last summer, Shapeways and Hasbro launched a revenue-sharing initiative at the site SuperFanArt.com, which redirects to a page on Shapeways's main site aimed at the growing niche of adult fans of the 1980s children's animated series My Little Pony.

Shapeways's Weinberg says the Hasbro-Shapeways collaboration "takes what could be an adversarial relationship and turns it into a collaborative one." It's a model that rights holders and fan artists can use when dealing with intellectual property questions that arise with any emerging technology, not just with 3-D printing.

"We have all learned a lot from technological disruptions in the past decade or two, and the biggest lesson seems to be that the best response to your customers adopting a new technology is to meet them there," he says. More legal battles over 3-D printing are coming, and uglier ones too. Weinberg underscores the upside to all the fighting for the future of the technology as a whole. "The good news is that all of this is a sign of a growing industry, and is really only happening because we are starting to see success."

The Netherlands spent seven years and $202 million to digitize huge swaths of AV archives that most people will never see. Was it worth it?

Driving through the Dutch countryside near the town of Hilversum, I have an overwhelming feeling that the surrounding water will wash out the road, given that my car is almost level with it. So it's surprising that the Netherlands' main audiovisual archives at the Sound and Vision Institute reside in a multilevel underground structure here, ostensibly below sea level.

Sound and Vision, together with two other national institutions, finished digitizing the bulk of the Netherlands' audiovisual archives last year, for a cost of $202 million over seven years. The project ran smoothly and transparently, digitizing 138,932 hours of film and video, 310,566 hours of audio, and 2,418,872 photos.

When Brewster Kahle, digital librarian of the Internet Archive, spoke at a conference here in March to commemorate the project's end, he called it "world-class" and told the audience, "It is almost exactly what Google has spent on their whole Google Books project, digitizing 20 million books." (Google says it has digitized over 25 million books but declined to confirm exact cost figures for this story.)

For all of Sound and Vision's efforts, though, only 2.3% of its digitized archive is publicly available online. Schools and researchers are allowed to access 15% of the archive on Sound and Vision's website. For the rest, Sound and Vision's administrators have to ask the copyright holders' permission to release their clips outside of the building. Frequently, it involves making calls to several people, and sometimes they say no. "Maybe digital formats are exploitable, but how much are old newscasts worth?" says Tom de Smet, Sound and Vision's head archivist.

Among the world's memory institutions, Sound and Vision's seven-year mass digitization project was one of the largest of its kind at its start in 2007, but few can reap its benefits today. Even as critics pressure the United States Library of Congress to digitize its collections, the Library must also solicit complex licensing agreements with corporate rights holders to release the files online.

"Record labels would want to make sure that we weren't putting up Louis Armstrong," says Gene DeAnna, acting chief of the recorded sound section at the Library of Congress's National Audio-Visual Conservation Center. "But maybe B-sides of Bing Crosby and things like that might be possible if they really had no plans to release them."

And even if memory institutions clear copyright hurdles—after countless hours of digital transformations and metadata documentation—they still need to make the sound and video clips findable. Their own online channels stand at odds with other online media platforms in engaging the public. YouTube will inevitably reach more eyes than a little-publicized Library of Congress sub-webpage. If going digital comes with so many caveats, is digitizing a country's entire audiovisual history ultimately more trouble than it's worth?

One Tiny Country's Digitized Archives

"It doesn't make sense to digitize everything," de Smet says in his office at Sound and Vision. "You have ask yourself, 'Who are you doing this for?'" Researchers may be interested in a narrow set of media, while the public may prefer a skim of the archives.

"Honestly, only a little bit of the funding should go towards digitization and the rest, towards digital preservation," says de Smet.

De Smet, a 36-year-old Belgian, is part of a new crop of archivists and librarians who aim to capture digital media's impact on the culture as it's unfolding. He believes that digitization on its own won't bring memory institutions into modernity. Rather, giant leaps of innovation will come from refining the methods used for preserving the digital files.

Now, Sound and Vision is negotiating broad licensing agreements with Dutch broadcasting companies, among others. One possible scenario under consideration is that Sound and Vision would be able to release a television show online 25 years after its initial broadcast. That way, neither party would have to dispute each creative work, file by file, saving time and administrative resources.

Sound and Vision's collection of television shows, newsreels, radio shows, music, and commercials dates back to 1898. Around 90% of the collections come from the Nederlandse Publieke Omroep, the national broadcasting agency. The other 10% comprises media of the institute's choosing. It even houses all of the Dutch royal family's audiovisual memories, but those will invariably stay private.

YouTube's "Media Memory"

Any new technology that better preserves and increases public access to these audio and video materials should aim to fulfill the greater mission of any national audiovisual archive: to be the "media memory" of the country. But what happens when YouTube pretty much serves that purpose today?

"By the time the project got underway, YouTube became immensely popular," de Smet says. It became clear to him that Sound and Vision's website was not the best way to present the available clips to the public.

"Collecting everything in one place online, it's a very linear way of thinking," de Smet says as he opens up a few tabs in his browser. He shows me Wikipedia, telling me that Sound and Vision is one of the biggest contributors of video to Wikimedia Commons.

To keep on top of the media culture, which increasingly happens digitally and online, de Smet sees an archive's role in society changing. Instead of centralizing the selection decisions in house, he and his team have struck up direct relationships with the creative community. Indeed, Sound and Vision has eliminated all of its curators and now trusts the community to curate its media memory. "Archives need to be at the start of the creative process," de Smet says.

Sony Steps Up

Gene DeAnna, from the Library of Congress's National Audio-Visual Conservation Center, launched the National Jukebox web app four years ago. For one year, throughout 2010, he and his team digitized 10,000 78 rpm music records, recorded between 1901 and 1925. All of those files are available online, thanks to a gratis license that Sony Music Entertainment gave to the Library of Congress.

"Sony has stepped up a lot. Others could do a lot more," DeAnna tells me in a telephone interview. He hopes to augment the National Jukebox this year with more Sony recordings, but plans to make material from other record companies available through the app are much further off.

"I have no interest in competing with record labels that are interested in reissuing historical recordings. To me, that's access," says DeAnna. Still, he sees more value in educating the public through the material—both the recorded media and the liner notes—than in trying to make money from the work of bygone artists.

Only a fraction of the Library's collections has been digitized. For perspective, the Library of Congress, the largest library in the world, acquires up to 100,000 audio artifacts per year but can only digitize around 15,000 recordings annually at capacity.

Professional Archives Vs. Consumer Platforms

The LOC reserves a lot of online real estate for the media that it can put on the Internet. Sites like American Memory, the Performing Arts Encyclopedia, and the American Folklife Center all reside under the larger Library of Congress property. The newly launched American Archive of Public Broadcasting, a collaboration with public radio broadcasters across the country, has a standalone site. Although the Library of Congress has a YouTube channel, it only uses it to disseminate video from its ongoings and conferences.

No matter which online content platform is used to host creative media, DeAnna stresses how important it is for the library to continue professionally archiving today's digital recordings.

"We are the de facto national library, so there's a sense of security there, and there's a long history of quality, data integrity, and cataloging. More longevity, more commitment to posterity than having content on a website. No matter how well established that website is, they just don't have capabilities that the Library of Congress does," says DeAnna.

Amazingly, the Library doesn't yet have a way to receive born-digital work as digital files. The record industry burns digital material onto recordable CDs, sends it to the copyright office for copyrighting, and the Library then rips the digital work from the CD-Rs to archive them on its servers. DeAnna says it is working with the copyright office, which is a department within the library, to implement a digital repository pilot program.

"It's a challenge with so much being produced to identify things that aren't copyrighted. Every year we find labels that haven't deposited, and then we ask the copyright office to write a letter to get the copies for us," DeAnna says.

A Global Effort

At the end of July, the Associated Press and British Movietone released 550,000 clips of historical footage to YouTube. More countries are undertaking or planning similar digitization projects as VHS tapes, audiocassettes, film, and other aging media carriers become unsalvageable in the next couple of decades. One can only hope that that work would also appear on the Internet.

The Library of Congress's collections comprise more than 450 languages, representing most of the world's cultures, so its preservation efforts won't only affect America's ability to remember its cultural past. With every innovation in communications—first the radio, then the television, and now the Internet—a country's media memory has increasingly spilled outside of its borders. Digitization will only speed up this cultural spread.

"There is no one single place that can serve the world's creative output. The more we can collaborate, nationally and internationally, the more successful we're going to be," says DeAnna.

As I leave Sound and Vision's in-house museum with de Smet, I see a blown-up print of Gene Simmons on a floor-to-ceiling wall montage. I ask de Smet what Simmons and his band Kiss mean to the Dutch media memory.

"When you grow up with it at the top of the charts, who's to say that it doesn't constitute a part of our national collective memory?" de Smet says.

A yoga startup increases its reach and looks to grab hold of the $60 billion digital health market. Is the concept a stretch?

During a recent class on the live streaming yoga app, Yoogaia, I started my computer's camera and touched my toes.

After a few poses, the teacher said, "Good job, Tina." Startled by having been addressed by name and feeling self-conscious, I smiled but then realized no one else but the teacher could see me. This is the maximum amount of interaction I prefer in yoga classes.

I seem to fit a couple of the typical user profiles of Yoogaia, whose classes I've been virtually attending over the last week. As a new-ish parent, I've found it difficult to find a good time to leave the house to go to the gym, and I travel frequently, making it hard to justify a monthly gym membership some months. And although I've never downloaded another interactive live streaming app before, like Meerkat or Periscope, I decided to give this one a try.

Helsinki-based startup Yoogaia (pronounced YO-guy-ah), a play on the Finnish word for yoga and the mythological Greek goddess, has a roster of qualified yoga instructors who lead a variety of classes in English, German, and Finnish, live streamed from each of Yoogaia's three studios in London, Helsinki, and Hong Kong. Covering three time zones, the daily live class schedule spans six to 20 hours of the day, with monthly pricing ranging from $11, for a yearlong subscription, to $22, for a single month. Users can choose to activate their microphones and video cameras, letting the instructor see and hear them in real-time. People who don't have time or prefer not to go to the gym can practice yoga, pilates, barre, and meditation classes wherever they have a connected device.

The question of who will adopt interactive live streaming technology is important, given that the latest such apps on the market—Meerkat and Periscope—are competing for users, but thus far have yet to implement monetization models. In contrast, Yoogaia has tapped into the $60 billion digital health market by leveraging live streaming technology to offer a product that lies somewhere between physically going to the gym and watching a prerecorded workout at home.

During his hectic career as an art and creative director for several large advertising and design agencies in Helsinki and nurturing an ever-growing family, Mikko Petaja found it increasingly difficult to make it to the yoga studio for his early-morning classes. That motivated him enough to found Yoogaia almost two years ago, where he is now the CEO. Since then, he has relocated to London, where he grows and supports Yoogaia's business.

Yoogaia has rapidly scaled up in several different markets since 2013. Originally starting out in Helsinki, Yoogaia opened its London and Hong Kong studios in 2014. In August, it received a $3 million seed round from Sanoma Ventures, Nokia Growth Partners, Inventure, and Point Nine Capital to support its international expansion. Last month Yoogaia started to test out classes for the German market and rolled out its iOS app.

Petaja tells me that Yoogaia will have over 100,000 users by the end of this year, and its revenue grows by 25% each month. There are 15 people on Yoogaia's team and 50 teachers at its three studios.

Yoogaia's plans to formally enter the U.S. market are longer term, and generally refining the product is a top priority. "We're less than two years old, but we have a global audience. There are a lot of things that we are still experimenting with," says Petaja.

Even without a physical studio in the U.S., more than 25% of Yoogaia's users come from the U.S. (Last week, that number climbed to 29%.) Anyone in the world can sign up for and use Yoogaia. "[The U.S. is] where a lot of our gold comes from," Petaja says. "The awareness of consumers in that market is in some ways higher in regards to using online services and potential online fitness solutions."

There is one drawback to using Yoogaia in U.S. time zones, however. The titles of some of the classes don't make sense (and the content of the classes may be off) when they are streamed live—"Good Night Yoga" is broadcast at 7:30 p.m. from London but streams at 2:30 p.m. in New York. Of course, most of Yoogaia's classes aren't meant for specific times of day, and all of them are archived on the site for people who miss the live sessions.

The company is also trying to better understand who its users are and refine how it delivers its service. One aspect is keeping classes small, with an ideal cap at 60 participants. "For the teacher, there's a limit to the number of students that he or she can instruct properly," Petaja says, alluding to the users' camera feature. "We may not want to have that many people, but from a technology point of view, we can scale up pretty quickly."

Yoogaia is not alone in experimenting with the live streaming fitness video market. (Just Google "live streaming yoga classes" for proof.) New York City-based Peloton, a spinning studio and stationary bike retailer, has an app that it uses to stream live and video-on-demand classes to thousands of users across 14 countries from its New York studio. Tulsa-based Live Streaming Fitness offers a broad range of fitness classes from instructors' homes or studios. On the social live streaming video app Meerkat, a few users specialize in fitness classes. None of these services offer video communication from the viewers to the instructors.

Crunch Fitness, the national fitness gym chain, live streams a class each Wednesday on Meerkat as well. In addition to the Meerkat stream, Crunch also runs a larger and decidedly more formal VOD service called Crunch Live, where subscribers access prerecorded workouts that are updated throughout the week.

Crunch once considered doing live streams of their physical classes but thought it would be too burdensome to the instructors. "We kind of made a calculated decision to go with prerecorded because we wanted the user to be the star and feel like someone was working with them directly," says Christina DeGuardi, Crunch's senior vice president of branding, marketing, and communications. "The teacher's attention would have been divided."

There are other business cases to be made out of busy consumers' demand for yoga experiences. Executive lounges at London's Heathrow Airport started offering sporadic yoga classes last year and now offer a special yoga room to elite travelers, a trend that includes airports in Chicago, San Francisco, Dallas-Fort Worth, and Burlington, Vermont. Hotels have been offering yoga classes to their guests, too.

Hyatt, meanwhile, once ran a yoga video-on-demand program called YogaAway, where guests could follow classes on their hotel rooms' televisions. The hotel chain eventually cancelled YogaAway because few guests were using it; this is one indication that Yoogaia's growth may hit some speed bumps. Not everyone, after all, wants to engage in something spirituality-leaning like yoga in a decidedly nonspiritual position: alone in bed, with a smartphone. And some folks need the motivation that comes from a room full of real-life, twisting humans to keep up the program.

Others, as Yoogaia's growth indicates, do not.

One night, after a particularly long day, I curled up in my bed with my phone and streamed one of Yoogaia's prerecorded meditation sessions from an earlier live stream.

As the instructor spoke calmly into the camera, she assumed her students were seated, but I reveled in the fact that I was lying down with the blankets over my head. I would have been mortified if someone in a physical class would have seen me in that somewhat childish position. But even if the on-demand option fit my life at that particular moment, I felt a twinge of guilt for having missed the live version from earlier that day. I made a mental note to attend the next live stream, but that seemed like a familiar tack from the time when I vowed to use my gym membership more often.

Rapid prototyping has graduated to on-demand production. Here's what that means for the future of manufacturing.

Jack Strauser's relationship with the Chinese manufacturing industry is incredibly tense. Strauser, founder and CEO of Florida-based company Dok Solution, believes a Chinese factory stole his electronic charging station designs and then offered them to a U.S. distributor, who now sells them.

"I don't have a million dollars to just throw around to defend my patent," says Strauser. "I'll spend every dime I have, but I have a lot of patents. It could financially ruin me."

Strauser is currently preparing to take the U.S. distributor to trial for patent infringement, with the claim that its main factory in Shenzhen poached his designs. But he fears that his small company may not be able to fight back against his much larger opponent. It's a David and Goliath story that, in many ways, reflects the greater mass production economy.

Reports about the faltering Chinese economy, the world's leader in mass production, bring up questions over whether tectonic shifts are coming to transform the manufacturing sector. And given the sharp increase in global hardware startups since 2012, it remains to be seen if small businesses, like Strauser's, have applied enough pressure yet on the manufacturing world to favor their interests. Rapid prototyping tools—namely desktop 3-D printers—have bolstered small businesses' paths to market in recent years. While China's manufacturing activity declines, is there an opportunity for 3-D printing to transform manufacturing even further?

3-D's Limits

"Digital printing is excellent, for me, to put my invention in a CAD design and have it made so I can get the samples. That's where it saves money in the United States," says Strauser, who has over 25 years of experience in manufacturing.

Strauser's charging docks are good candidates for 3-D printing. Reduced to their components, his docks comprise a few pieces of plastic and some electronics. A 3-D printer could handle everything except the assembly and the electronic components—the resistors, capacitors, and inductors—which would need to be ordered separately, from China or elsewhere.

Yet Strauser's production lines will remain firmly planted in Chinese factories. A 3-D printer wouldn't be able to handle the volumes of plastic Strauser needs to meet his clients' volume demands within an acceptable timeframe and with a good quality standard; his distributors ask for a 2,000-item minimum. Mass production, he says, is the only method that can accommodate that volume at a reasonable price point, so he's staying in China for that, choosing to struggle to keep his business afloat in the sea of factories and companies there.

For now, looks like the pace of change in the consumer electronics industry will be glacial erosion, rather than seismic. "There's no device that you're using today that can be 3-D printed to the standard you're going to accept as the consumer," says Liam Casey, founder and CEO of PCH International, a company that creates manufacturing and retail solutions for growing businesses. One of its main activities is linking its clients to manufacturers in China. "None of the companies we work with would accept the quality of products that comes off a 3-D printer."

Outside of the prototyping phase, only certain industries, like medical and aerospace, have been able to derive value from 3-D printed products; they thrive on customized, one-off parts. Consumer electronics businesses, on the other hand, are confronting 3-D printing's limits now that they are transitioning from prototyping and beginning to use the technology for on-demand production.

A Brooklyn startup called BotFactory sells a tool that enables inventors to produce printed circuit boards on demand. Though it doesn't exclusively use 3-D printing, BotFactory's on-demand production principle is the same: The inventor loads a design, and the system instantly starts making the product. First, a user loads her PCB design into a web-based form. Then, a machine prints the connections onto any flat surface with conductive ink. Finally, a robotic arm automatically places the electronic components into the appropriate spots on the board.

But BotFactory acknowledges that its system is only suitable for prototypes and low volumes; printing anything larger than 1,000 pieces in a reasonable amount of time would require several of BotFactory's $2,999 machines working under human supervision. For many smaller companies, outsourcing to China for high-volume projects is worth the time and cost savings.

Build Local?

Proponents of the local manufacturing movement say that production trends are firmly pointed away from China's mass production economy. Startups want to produce their products close to their designers, cut down on the amount of inventory they store, and reduce delivery times.

"It's going to be on demand, it's going to be local, distribution is going to be free, basically," says Brian Garret, cofounder of 3D Hubs, a Dutch startup that connects people who want to 3-D print with local 3-D printers within its online network. In 2013, 3D Hubs's network had just nine printers. This year, the network surpassed 22,000 printers, giving 1 billion people access to 3-D printers within 10 miles of their homes. It is the world's largest network of 3-D printers.

Garret has his eyes set on upsetting global, commercial delivery times, aiming to attract companies with Asian manufacturing operations, like Nike and Ikea, as clients. Meanwhile, in February, Amazon filed a patent for delivery trucks that would be outfitted with 3-D printers, yet another move in the company's quest to realize on-demand delivery perfection. But it's not clear exactly when these projects will materialize.

A startup called CloudDDM built its 3-D printer farm within one of UPS's logistics hubs in Louisville, Kentucky. Mitch Free, CloudDDM's CEO, says he can print until 1:00 A.M. everyday and have parts sent to any client in the United States by 10:00 a.m. using UPS's priority service. He now has 100 printers that run with little supervision.

"We are doing low-volume production, creating high-quality parts, with high-grade thermoplastics," Free says. Any high-volume orders are out of his capacity range, but he claims that's not an issue; businesses want to stock fewer items on their shelves. "No one wants to keep inventory; when they need more they can just order more," Free says.

Beyond Printing

While 3-D printing seems to be stalling in the consumer electronics industry, another technology is gaining traction: digital manufacturing. A central component for on-demand production, digital manufacturing uses the tools of big data, sensors, and communication technology to automate as much of the manufacturing process as possible.

This past May, the Chinese government announced its "Made in China 2025" plan, which would modernize its manufacturing sector with digital technology to compete with emerging trends in competing countries. Casey says that China's manufacturing practices have already been evolving from the black box reputation it earned in the 1980s. The Chinese manufacturers that PCH International works with have made their production processes more transparent and are enthusiastic about supporting their clients' brands and quality standards.

And digital manufacturing's prevalence in the U.S. small business space is also growing. This year, Phoenix-based Local Motors 3-D printed an entire car chassis in just 44 hours, for its electric car, the Strati. Local Motors hopes to clear U.S. auto regulations in 2016 to produce the car for consumers at its expanding web of "microfactories," hyperlocal spaces that integrate all of the aspects of design, manufacturing, and retail. Etsy, meanwhile, launched its Etsy Manufacturing business in September to match Etsy sellers to small manufacturers.

And Brooklyn's Maker's Row specializes in digital manufacturing to facilitate startups' transition to the manufacturing phase. "You can develop your product online, you can pull a supply chain together online, and you can correspond with everyone," says Matthew Burnett, Maker's Row's CEO and cofounder.

For now, Dok's Strauser still goes to China five or six times per year. Lately, however, he has had trouble finding reliable factories to work with, witnessing firsthand how they are struggling to stay open, amid pressure to increase domestic wages and accommodate labor law changes. Over the years, he has moved from factory to factory, juggling three to four contractors at a time. If he had the money, he says, he would manufacture as much as he could in the United States. And 3-D printing wouldn't be a part of the process.

Will Turnage is SVP of technology at the R/GA agency. And cocreator of the algorithm-powered cooking app, My Robotic Kitchen.

Will Turnage loves food. But he may love wielding data and technology to master cooking even more.

So for the last seven years, Turnage, senior vice president of technology for the R/GA advertising agency in New York, has been on a food app bender.

For one of Turnage's early app projects, he looked close to home: to the dinner parties he and a friend, Mike Lee, founder of the supper club Studiofeast, enjoyed hosting. And the duo realized that planning one is difficult and time-consuming to manage. Everything from scaling recipes to accommodating guests' food preferences to optimizing cooking steps to knowing when to start the apéritif was much more involved than a single busy person could handle on a Saturday afternoon.

That fact inspired Turnage and Lee to develop the algorithm behind the My Robotic Kitchen app, to make all of those planning decisions instantaneously. In 2012, Turnage and Lee threw a dinner party with the software and presented a demo of it at SXSW.

My Robotic Kitchen and Turnage's other extracurricular food app projects have earned him the unofficial title of R/GA's "food guy." Now, Turnage consults on all of the food-related accounts, and not just in a technical capacity. "I had suddenly gained this reputation internally for a content area, and that cuts across all disciplines," Turnage says.

Not only has his outside work opened a lot of doors for Turnage at R/GA over the years, but his work at R/GA has helped him broaden his food hobbies as well. While R/GA has a diverse set of clients, the agency has at least one finger in the food space, mentoring connected kitchen startups and co-sponsoring food and technology events. "It sort of creates this happy, symbiotic feedback loop," he says.

When Turnage first started at R/GA almost 10 years ago ("That's like four lifetimes in the tech world"), he worked on the technical side of the Nokia account, when Nokia's feature phones ruled what would become the smartphone market. His food obsession drove his early knowledge of the mobile market.

"All people were doing was scanning pages of cookbooks and putting them on a website. It wasn't interactive at all. At the same time, the mobile world was new, and I wanted to learn something about it," Turnage says.

At the time, Turnage had been following the food writer Michael Ruhlman, a James Beard award winner who has appeared on the Food Network's The Next Iron Chef television series, and knew that Ruhlman had recently published a new book. Turnage cold-called Ruhlman to see if he was interested in developing a companion mobile app for it. The pair ended up creating and releasing a couple of food apps together: Ratio and Baking Bread Basics.

Turnage talks about cooking apps like a human blender: Once you get him going, you have to wait for the right moment to release the button. As Turnage finishes telling me about an R/GA-sponsored hackathon with Food+Tech Connect to innovate in the meat industry, he mentions that he once created a sensor system for monitoring the meat-smoking process. When he showed it to a chef friend of his, the friend realized it would be perfect at managing the soy sauce fermentation process. New York's Momofuku Lab ended up using Turnage's system to monitor its soy sauces.

While Turnage and Lee were developing My Robotic Kitchen, there came a point when they realized the software couldn't dissect and resynthesize the prose of several recipes' directions simultaneously. The software worked well on a small level, accommodating a few recipes at a time. Perfecting it would have required Turnage and Lee to outsource or scale back their day jobs. Says Turnage: "We had to decide: Was this a business?"

Last year, Turnage put My Robotic Kitchen in his virtual archive and blogged about it—among other stalled food app projects—on Medium. Collectively, he learned that these apps' business models weren't strong enough to sustain them.

But Turnage doesn't see these unfinished projects as failures. "I think it's really important to make things to really succeed at that senior level, where you can work with creatives to help them understand how to build things."

Turnage thrives at R/GA, and in the hours when he's not there, developing his food apps is just one of this side projects. But now that he's filed several of his ideas away, he has been making more room for other activities. He is an avid swimmer and a member of a synchronized swimming team. And yes, the team has a food-related name: they are the Brooklyn Peaches.

Model Ashley Graham is turning her lingerie line into a cause. This is good news for Lena Dunham, Serena Williams, and women everywhere.

Walking triumphantly down the Tommy Hilfiger runway at the end of a show during New York Fashion Week in September while holding hands with her fellow models, Gigi Hadid looked happy and confident.

But after she posted a photo of the moment on Instagram, she received thousands of comments—and many of them were negative assessments of her body, some saying Hadid was too thin, others saying she was too large. The comments troubled Hadid, and she wrote a public response in defense of herself, which she also posted on Instagram.

"Your mean comments don't make me want to change my body," Hadid wrote, in part. "They don't make me want to say no to the designers that ask me to be in their shows, and they definitely don't change the designers' opinions of me."

The instantaneous criticize-and-defend repartee over a woman's appearance on social media has given broader reach to an unfortunate trend: body shaming. On Instagram, Twitter, and Facebook, women are body-shamed by people who post negative comments about their pictures. Women like Hadid, who are often celebrated for their bodies, are not immune to this trend. Star athlete Serena Williams is the subject of body shaming during each tennis season. Writer and director Lena Dunham recently said she was no longer looking at Twitter after her own body-image-related dust-up with commenters.

"Even if you think, 'Oh I can read, like, 10 mentions that say I should be stoned to death' and kind of, like, laugh and move on, that's verbal abuse," Dunham told the journalist Kara Swisher. "Those aren't words that should be directed at you, ever. And so, for me personally, it was safer to stop [using Twitter]."

While in some corners of the web, such abuse continues apace, in other industries we may be witnessing a turning of the tide. Now support for women like Dunham, and Williams, and anyone else who's experienced body shaming is coming from a corner of the retail clothing market not normally associated with progressive attitudes on body image: women's lingerie brands. Lane Bryant, who creates and sells clothing for curvy women sized 14 and up, promoted its lingerie line over the past year with two hashtags, #ImNoAngel and #PlusIsEqual. The first message is a direct critique of the generally unattainable slim Victoria's Secret Angel. And the second suggests women who wear larger sizes deserve fair representation in the fashion and media worlds.

More recently, model and body-positive advocate Ashley Graham began to promote her own lingerie line for curvy women with the hashtag #IAmSizeSexy. She had previously modeled in both of Lane Bryant's hashtag campaigns. In February, Graham became the first curvy model that Sports Illustrated featured in its annual swimsuit issue. There, she appeared in a full-page advertisement for the swimsuit brand SwimsuitsForAll's #CurvesInBikinis campaign. In September, Graham premiered her line with the #IAmSizeSexy hashtag at New York Fashion Week.

An Activist Is Born

Graham started modeling when she was 12 years old. Her unconventional shape automatically classified her as a plus-size model, a fashion industry term that gradually made her feel uncomfortable. Although Graham was naturally outspoken and confident, the constant scrutiny in the fashion industry eventually took a toll on her mental and physical health. So she decided to become an agent of change as an advocate for curvy women.

By 2012, when she was 25 years old, Graham had already been thinking about creating a lingerie line for her body shape. One day, while waiting on a photo-shoot set for the Canadian retailer Addition Elle, she spontaneously pitched the company's vice president for marketing, who was sitting nearby. The director asked her for details on how her designs would be different from other lines out there.

"I lifted up my shirt," says Graham. While pointing to different parts of the bra she was wearing that day, she explained where she would change the lace, trim, and fit. "It was literally one of those moments where I knew if I wouldn't have asked, I would never have known." Graham has been selling her line with Addition Elle since 2013.

When Graham created her #IAmSizeSexy hashtag, she knew she wanted it to become an empowering tool for women to use online. Curvy women already participate in vibrant communities online. Women propagate positive hashtags like #EffYourBeautyStandards and #HonorMyCurves through Twitter and Instagram, to stir discussion around curvy women and create a support network. "Whatever sort of hashtag you identify with, it's incredibly political," says Sarah Conley, a blogger and social media consultant. These hashtags, Conley says, help give women of all body shapes a collective voice to tell the fashion and media industries that they want to be respected and represented.

"The word 'size' had to be in it," says Graham. "It's brought a community together, especially for curvy women, who are supporting each other, lifting each other up. It's actually bringing a sense of feeling that you belong on social media, because you get to find your community when you have that hashtag."

Economics Of Scale

Curvy women have been marginalized in the fashion industry for decades. Established lingerie brands have traditionally not carried their lines in larger sizes because they perceive a low demand for a process that would require a high investment in design and manufacturing to accommodate more variable body shapes. Market research firm Plunkett Research estimates 67% of women wear size 14 to 34, which the fashion world categorizes as plus-sized. Yet only 9% of the $190 billion spent annually on clothes is spent on plus-size clothes. All of which means that the demand is there, and retailers have an opportunity to better serve the plus-size market.

Some lingerie lines are helping women build online communities to celebrate their bodies. A brand called Curvy Couture runs the campaign "Curvy Girl of the Month," wherein a different customer features prominently on the company's website each month. Another brand, Jewel Toned, which is not exclusively for curvy women, runs a similar campaign on its Instagram feed and blog and calls their featured customers "bombshells." Aerie launched its #AerieREAL campaign last year, showing women with untouched photos in its ads.

"It's such a body-conscious, form-fitting, emotional process, both positive and negative, for women to buy intimate apparel," says Rachael McCrary, founder of Jewel Toned, a lingerie and shapewear company. McCrary believes lingerie brands should feel some responsibility for caring for their customers' concerns.

Graham, meanwhile, released her holiday line this month at Addition Elle and at Nordstrom in the United States. Last year her line grossed $1.6 million. And Graham expects to surpass that figure this year. "I personally believe that the increase in sales proudly shows that our message of body acceptance and fashion equality is definitely being heard," she says.

Graham's show at New York Fashion Week was not only a way for her to show off her designs. It was a forum for celebrating women's bodies. Graham radiated confidence as she modeled her designs alongside her show's models. And when she posted photos of those moments on Instagram, they added to the positive movement that she has helped to inspire.

Benjamin Gaulon, a veteran digital artist, used his connections to cultivate the teaching faculty at the New School's Parisian campus.

Since the early 2000s, French artist Benjamin Gaulon has built his career on revealing the faults in some of today's most popular consumer devices. He and a friend once wrote display-scrambling software that numerous people have downloaded onto showroom computers inside Apple Stores while filming customers' bewildered reactions. In another project, Gaulon ceaselessly trawled eBay for defunct Amazon Kindles, signed the backs of them, and put them back for sale on Amazon. Though his projects border on the rebellious, Gaulon also manages to celebrate technology's possibilities.

So as Gaulon's wayward niche started to evolve, he gradually nurtured a parallel academic career in teaching technology that institutions of higher education have come to respect. Today, Gaulon is the program director for art, media, and technology at Parsons Paris, The New School's longstanding art and design school in Paris. Since joining Parsons Paris in 2013, Gaulon has attracted globally respected artists as adjunct faculty and helped institute degree programs in technology that distinguish Parsons Paris from its sister institution in New York, all while continuing his own exploits in the art world. It seems that for Gaulon, educating future creators serves the same purpose as his subversive artwork does, to sustain a dialogue about the significance of today's technology.

Increasingly more American universities are opening international branch campuses that teach the home university's curriculum in the foreign country and confer degrees from the home institution. The Cross-Border Education Research Team reports that as of this summer, U.S. universities have opened 81 international branch campuses, far more than any other branch campus exporters in the U.K., Russia, France, and Australia. Globalization is rapidly expanding the educational market, which IBIS Capital estimated as a $4.4 trillion market in 2013.

Yet Parsons opened its school in Paris decades before other universities started to move their campuses abroad. When it opened its art school there in 1921, it became the first American college in any discipline to establish a study-abroad program in Paris. Parsons Paris continues to make a unique proposition on the educational market with its technology-infused curriculum from one center for art and fashion to another. It has three bachelor's and four master's programs, and since it reopened in 2013 after a three-year reorganization, the school has grown to 150 students across all programs. Susan Taylor-Leduc, dean of Parsons Paris, says she expects to admit 50 more students next year.

The things that make Parsons Paris special became tight constraints on Gaulon's quest to build his team of adjunct faculty for his specific programs. He had high standards for finding local talent that not only embraced The New School's New York essence, but could also leverage Europe's cultural resources.

"In France, to find people who can teach well—and in English—is hard to find. Most of them already work with me," Gaulon says. He knew that the search for excellent academic talent was global, so he reached out to his network across Europe and North America.

The people Gaulon has helped hire at Parsons Paris are quasi-luminaries in the digital art world. Digital artist Chris Sugrue coteaches a master's studio class with Gaulon. And Alessandro Ludovico, chief editor of the new media arts criticism magazine Neural, helps mentor Gaulon's master's thesis seminars. Next semester, the globally renowned digital artist Evan Roth, who also happens to be Gaulon's studio mate, will teach a course in Gaulon's BFA program called "Internet Landscape" for the first time.

Both Sugrue and Roth are graduates of Parsons's MFA in Design and Technology program in New York. "It's super-nice to have them here to get a bit of the spirit of New York in Paris, but with our own flavor and old style," Gaulon says.

He has gathered a multinational group of programmers, photographers, sound performers, publishers, DIY electronics people, and what he calls "urban hackers." Out of Parsons Paris's more than 50 faculty members across its four main programs—fashion, design history and criticism, and design strategy comprise the other three—Gaulon has chosen 13.

Gaulon spent more than a decade away from his native France to earn his MFA in the Netherlands and then explore the creative scene in Dublin. While teaching in an arts graduate program at the National College of Art and Design there, he heard about an opening at Parsons Paris through some friends. He says being French strengthened his application, but by then, his professional network was almost exclusively non-French. Gaulon's global network ended up being what Parsons was looking for.

At Parsons Paris, Gaulon created his bachelor's program, the BFA in Art, Media, and Technology, "basically from scratch." It is only available in Paris and not at Parsons's New York campus. And he brought Parsons New York's popular MFA in Design and Technology to Paris last year.

"People are creating their own software and hardware and really developing new types of technologies, researching what it means to work with technology today on a theoretical and practical level," Gaulon says of his students. The programs teach electronic programming and creative research around art and media technology.

The curricula that Gaulon helped design are exceptional to what's on offer at virtually all French art schools and most similar schools internationally. The competition to Parsons Paris's art, media, and technology programs, he says, is largely global, counting programs at New York University's Tisch School of Arts and the Royal College of Arts in London as competitors.

Last year, the first pair of master's students to finish Gaulon's new MFA program showed their graduation projects at a dedicated digital art gallery in Paris. Both of the students were women. He recognizes that women dominate his classes and strives to have his faculty reflect that gender makeup.

Gaulon is constantly struggling to find a stationary point in the interplay he sees between technology and art. He wants his fellow faculty and students to also try to interpret this exchange. Sometimes, he says, he's not sure that there's one teacher out there that can fit his singular vision. Perhaps that person doesn't exist yet.

He encourages his students to keep an open mind when learning new concepts of art and technology in his classes. "Some of my students say, 'I think I should be focusing on doing something that I can do as a job.' I tell them that this is the time to really ask the big questions."

This month, Gaulon's 2.4 GHz project will start showing at the ZKM Center for Art and Media. Later this month, he'll collaborate with art colleagues at the School of the Art Institute in Chicago. And in November, he will show at the Utah Museum for Contemporary Art in Salt Lake City. Gaulon's students in Paris aren't the only ones who can learn something from his hacker ethic.

Orr Shtuhl of design agency Blenderbox teaches classes on pairing cheese and beer. Company morale: Meet your master.

Only ask Orr Shtuhl about cheese, beer, and cocktails if you have a few minutes to spare.

The user experience director at the interactive design agency Blenderbox, Shtuhl happens to know an awful lot about cheese pairing. Ask him about cheese and beer or cheese and cocktails, and he can give an entire lesson on the topic. In fact, he gives regular classes at cheese retailers in Manhattan and Brooklyn. (He gets a little insecure when it comes to pairing cheese and wine.) His students are similarly perplexed at the beginning of his classes, when they set out to understand the parallels between non-wine beverages and cheese.

"The spectrum of flavors is wider with beer," says Shtuhl, "but [wine and beer] are both equally complex."

Just as with his surprising cheese pairings, Shtuhl has learned to appreciate how his two seemingly disparate vocations are actually more complementary than he first realized. It wasn't until recently that he understood how his mode of teaching is as carefully thought-out as his web-design work in his day job at Blenderbox.

Shtuhl's teaching skills enhance his frame of thinking at work every day. "I always approach [UX design] as education. You need to think about what the user's level of knowledge is when they get here, and what the level of knowledge is that you want to take them to."

As the director of user experience, Shtuhl oversees how clients will perceive and interact with all of Blenderbox's final products. And at his pairing classes, he makes sure that the educational elements harmonize with the evening's bright atmosphere.

"In both cases, it's important to start with not just: 'What is my favorite thing?' or 'What do I want to tell people about?' but also, 'What does the person want to know?'" says Shtuhl. But he adds, "The first goal is to always have fun."

For years, Shtuhl lived a kind of dual life as a food journalist and web design expert. Before he came to Blenderbox in 2011, he reported on beer trends for the Washington City Paper in Washington, D.C., while working as a web editor for a think tank during the day. He even ran an illicit speakeasy from the house he lived in in D.C.

While Shtuhl was climbing the organizational ladder at Blenderbox, he wrote and published a book about cocktails, complete with illustrations. Now, he's grown into both roles, as an expert of beverages and web design.

"It's basically information architecture. When I give classes, the concepts are very structured, in my mind at least, and I start everything with bullets," Shtuhl says.

He fits nicely into Blenderbox's epicurean culture. The client roster reads almost like a gastronomic retailers guide: the Brooklyn Brewery, the Brooklyn Kitchen, the artisanal food store D'Artagnan, and others. And office huddles are known to include a big brewing pot and intra-office brewing sessions. This boosts office morale, while allowing Blenderbox to practice packaging design.

Blenderbox cofounder Jason Jeffries, meanwhile, runs one of the venues where Shtuhl teaches, the Bedford Cheese Shop—when he isn't programming in the office, that is.

This month, Shtuhl launched his own beer-tour business, a project that will use skills from both his informal and formal jobs. This gave him yet more experience building a website and developing a brand concept.

"That's always been my life," Shtuhl says. "Whenever I have a spare moment, I think of some other side project to do."

Whatever his next project is, don't be surprised if it's spirited.

Look for them in Boston in April.

In an attempt to appeal to design-conscious consumers, major athletic shoe brands like Nike and Adidas have been investing heavily in 3-D printing. Now, New Balance is aiming to beat both Nike and Adidas to the finish line.

Today, the company announces a new running shoe model that incorporates a 3-D printed midsole. The shoe will be available in April in Boston as a limited edition collection, with the release timed to coincide with the Boston Marathon. New Balance will sell "a couple hundred" of them in 2016 and intends to expand their retail operations to more global metropolitan locations thereafter.

The design is an update to New Balance's Fresh Foam Zante model. But instead of gluing a traditional foam midsole into the shoe, the new shoe will incorporate a 3-D printed midsole. The upper of the shoe will still be composed of traditional cloth. Though New Balance has not released exact price figures, it says the market price will be higher than New Balance's regular range of running shoes and hopes they will appeal to serious amateur runners interested in innovative footwear.

New Balance expects to offer consumers the opportunity to customize their 3-D printed midsoles as part of a trial program, in select retail locations, in 2017.

Shoe brands are competing to achieve a lighter, more quickly made shoe, and they are using 3-D printing to do it. In October, Adidas announced it had created a 3-D printed midsole prototype as part of its new Futurecraft 3D project, loosely committing to bringing a limited quantity to market by next summer. At the GeekWire Summit in Seattle, Nike COO Eric Sprunk suggested that Nike is accelerating its own 3-D printing efforts.

3-D printed midsoles have generally been too heavy and rigid to benefit athletes, but New Balance says it has found a material that achieves an acceptable mix of flexibility and weight for performance wear. This year, New Balance partnered with South Carolina-based 3D Systems to incorporate one of the company's' new elastomer materials into its printing process.

The resulting elastomer-made midsole is less massive than New Balance's previous printed midsole prototypes, and light enough that New Balance feels ready to introduce it to the marketplace. The printed midsole's current design is about 40 grams (1.4 ounces) heavier than a full-length, foam-only midsole, but the New Balance team plans to make improvements between now and April.

"In design, we have been erring on the side of durability and still have opportunity to optimize for weight," says Katherine Petrecca, New Balance's general manager for studio innovation.

New Balance's 3-D printed midsoles reflect improvements in the company's design method, leveraging the convenience of 3-D printing. Foam insoles are typically created with injection molding, which wastes unused material and requires continually purchasing custom molds at a few thousand dollars apiece, each time.

New Balance uses a specific manufacturing technique called selective laser sintering to 3-D print the new midsoles, building up layers of material instead of cutting away scraps, without the use of a mold. New Balance's SLS printing method uses lasers to fuse the proprietary elastomer powder into a cohesive material for the midsole.

"With injection molding, there are significant restrictions in how you can design a part or how you can engineer variable properties into a single part. And that's what's really attractive about SLS printing," says Petrecca. "It really opens up design capabilities in being able to design a part down to the millimeter and opens the door to performance customization."

New Balance has been experimenting with 3-D printing since 2013, when it started to produce 3-D printed, cushioned spike plates for competitive track athletes. A spike plate, which is different than a midsole, accommodates spikes or cleats that dig into the racing surface. New Balance has since fitted professional soccer and baseball players with these 3-D printed spike plates. They are currently only available for a select group of professional athletes, called Team New Balance.

New Balance customizes the spike plates to the performance characteristics of each athlete, which may, in turn, provide a competitive advantage. In June 2014, professional track athlete Kim Conley won the 10,000-meter event at the USATF Outdoor National Championships while wearing New Balance's 3-D printed spike plates.

But consumer customization is far off; it will require its own design overhaul before it's ready for market. The technology New Balance currently uses to measure an athlete's exact movements involves complicated sensor configurations and lengthy movement observation sessions with expert bio-mechanists.

New Balance anticipates its consumer customization process will de-emphasize extensive biomechanical measurements and involve only some of the individual's measured physical data, like his or her weight, with reported running preferences.

One startup claims to achieve total insole customization already. SOLS, a Manhattan-based startup who counts among its investors New York Knicks basketball player Carmelo Anthony, has a proprietary customization method that only requires feeding a few photographs of the wearer's foot into an algorithm. The algorithm estimates the 3-D shape of the foot, and the resulting 3-D printed insoles can be used in any type of shoe, not only in athletic footwear.

The athletic shoe industry recognizes that innovation outside of 3-D printing will translate into better shoes, too. This fall, Adidas launched its Sport Infinity program, which aims to create completely recyclable shoes as part of the company's greater effort to make manufacturing more sustainable. Earlier this month, Adidas released a new Futurecraft creation, a shoe whose upper is composed of a single piece of leather, made with digital milling. (It released 45 pairs to the public for one day on November 7, in New York, London, and Tokyo, but there are no publicly known plans to release the model to the wider public.) New Balance, like Adidas, is also exploring digital milling in its in-house prototypes and other manufacturing processes to enable more design improvements to both midsoles and uppers.

It isn't clear when any of these brands will produce 3-D printed footwear on the same levels of traditionally mass-produced product lines. They will have to wait until 3-D printing becomes less cost-prohibitive at higher production numbers. In general, once production levels of a printed product top 1,000 pieces, 3-D printing's costs start to outweigh its gains in time savings. At high quantity, producers default to mass production methods, like injection molding.

That the new, 3-D printed New Balance shoes will be available to the mass consumer represents a milestone for 3-D printing overall, says Petrecca. Whereas 3-D printed products are typically made from hard materials, New Balance believes it has found a unique method to print soft materials by combining materials innovation with plastics engineering.

"We've made some significant advancements in the performance of [3-D] printing, to create parts that are light and soft and flexible, and mimic and replicate what we are able to do with foam," Petrecca says. "And to create that in a package that's also durable enough for running use."

Amy Robinson, executive director at the startup EyeWire, is making neuroscience into a playground for the hot tech du jour.

Virtual reality is here, and brands of all stripes are embracing the tech. The New York Times and Google newly partnered to send more than 1 million cardboard VR viewers to Times subscribers at the beginning of November so they could watch the paper's first VR documentaries on a smartphone. Magic Leap published video of its augmented reality system online in October, causing more buzz around VR's potential. Even Tommy Hilfiger now offers VR sets to its in-store customers so they can watch its recent New York Fashion Week show.

But major brands aren't the only early adopters of the latest technologies. An emerging group of science startups is using hardware innovations to create useful applications out of basic scientific concepts. One such startup, EyeWire, crowdsources computing power to visually map out delicate cells, or neurons,in the brain—and packages it as an online video game at EyeWire.org. Though the players' 3-D visualizations primarily aid scientists, EyeWire's executive director, Amy Robinson, has been using the mapping data to experiment with virtual reality and other tools beyond gaming to improve everyone's understanding of the brain.

"The brain is this incredible place for data visualization," Robinson tells me by phone from EyeWire's Boston office. The billions of connections between neurons are enough to bring the latest tools of big-data science to their breaking points. That's why EyeWire crowdsources video-game players to map out its neurons; neither an artificial algorithm nor a normal computer could create the same 3-D maps. EyeWire's players, then, make up a kind of supercomputer with human intelligence.

The game comes loaded with 3-D images of several neurons. One neuron appears on the left part of the screen at all times; as the player becomes more adept at mapping, different neurons appear, their shapes increasing in complexity. On the right side of the screen is a 2-D cross-section of the neuron. Both images show where EyeWire's artificial intelligence algorithms have begun to color in the nerve cell; these shapes are shown in blue.

The player then spots gaps that the AI missed; those voids are shown in gray or yellow, depending on the level. When she clicks on the missing piece, it turns green. If she chooses a section that the game's algorithm definitely thinks does not belong to the selected neuron, the section turns red, and pop-ups urge her to deselect those pieces and start over. She can modulate the 2-D image she is playing with by scrolling through the 3-D image on the left. Once she fills in all the missing pieces, she moves to the next level. EyeWire intersperses the coloring challenges with periodic "Starburst Challenges" and "Happy Hours", where players can score exclusive points. All of the players' completed maps become part of a greater, collective neural "map."

Neuroscience is a field for limitless exploration and new discovery. The biggest technical revolutions of the last century—nuclear energy, computing power, and space exploration—all started with basic science research that had no immediate industrial use, but evolved into important industries with rewarding applications. Now, science advocates say neuroscience is the next revolution. By exploring the mysteries of the brain, budding technologies could benefit. These technologies will, likewise, enhance basic research. So EyeWire's Robinson is eager to venture into yet more unconventional tech projects—VR is just one of them.

Robinson's interest in human consciousness is as winding and tangled as the neural connections in the human brain. She wants to understand exactly why she exists, and everything from science to the canonical philosophical thinkers—Nietzsche, Seneca, and Voltaire—feeds her quest for knowledge. The number of neural connections in the brain amounts to 100 terabytes. Indeed, if Robinson's enthusiasm for neuroscience were converted to bytes, the Google Hangout she and I are having would crash.

"If you laid all the neurons in the brain end-to-end, they would go around the earth four times," Robinson says. Robinson puts her science communication skills to good use as a TEDx organizer and speaker; Forbes added her to its "30 Under 30" list this year in its games category.

Last year, neuroscientist Sebastian Seung and his team at MIT published an article in Nature that listed all of EyeWire's then 2,183 contributors as co-authors. Their mapping work led Seung's team to better understand the way humans see. Robinson is largely responsible for attracting those contributors to the game.

When Robinson joined EyeWire in 2012, the site, she says, was unattractive and had few players. Using her instinct for making connections with kindred creative spirits and her love for the latest design trends, Robinson brought in talented designers and drove EyeWire's creative operations. Under her leadership, EyeWire's software has become a model for scientific data visualization, having garnered top prizes from the science community. Its visualizations have even been featured in Times Square, and its player base has grown to 200,000.

Now, Robinson is uncovering new ways to understand neuroscience with hardware. In 2014, the White House invited Robinson and her team to contribute to a panel on modernizing academic research with 3-D printing. After putting their first 3-D image files onto the National Institute of Health's 3-D print exchange repository, a Canadian biology major named James Drake unexpectedly discovered his 3-D printer's limitations. The printer couldn't render one of the neuron's delicate structures well, so he refined the virtual model by hand. Robinson's team quickly went to work on an algorithm to automate thickening the 3-D models that other scientific labs could use, in collaboration with Drake.

In April, the EyeWire team uploaded their "Neuron Safari" VR application onto Oculus Rift's public repository, the first in a series of collaborations with the medical visualization company Indicated. But this wasn't EyeWire's first VR application. At TED2014, EyeWire set up a VR exhibit where visitors could don an Oculus Rift to virtually travel to the International Space Station and then tunnel through space-buoyed neurons. "The brain is so complicated that I think VR or AR are necessary for the future of mind mapping," Robinson says.

Many neuroscience professionals champion VR's prospects. "Virtual reality, for the human brain, is a new mode of information delivery," says Elkhonon Goldberg, a neuroscientist and chief scientific officer at Monclarity, a company that produces a brain training game called Brainwell.

"We're only beginning to ask questions of how the human brain will catch up to this totally new environment," says Goldberg. The Brainwell team is currently developing VR applications for its game.

EyeWire solely relies on public and private grants as a nonprofit startup. But more science startups are beginning to find venture capital funding, which could spawn EyeWire-like businesses. As government funds to support scientific research have cooled down, the venture capital world has been warming up to investment opportunities in tech-driven science. The Thiel Foundation's seed-stage fund, Breakout Labs, exclusively seeds science startups. NeuroLaunch, an Atlanta-based startup accelerator that invests exclusively in neuroscience startups, graduated its first two classes of startups this year. And since 2014, Y Combinator has been steadily increasing the number of science startups it funds. It even opened its own research lab to support scientific research from which its classes could benefit.

"Now that the startup ecosystem is developing more, and there's more people working on these deep technologies, these technologies will start to form an infrastructure that science labs, including neuroscience ones, can use to translate basic concepts into useful applications," says Andrew Wong, founder of a new science startup fund called Boundary Impact Ventures.

Robinson, however, plans to exercise EyeWire's independence from venture capital investment to keep exploring new avenues for neuroscience to mingle with emerging technology.

On a recent evening, 57 players were on EyeWire.org. I joined the English-language chat room on the left side of my screen. (The other available chat room is in Korean; EyeWire's largest grantor is the KT Corporation, South Korea's largest phone company.) On the right side of the screen, I could see today's top scorers, most of whom resided in Western Europe; one was from South Korea. The highest-ranking American was listed as number 12.

As more people play EyeWire, neuroscientists receive more 3-D data to support their research. And Robinson receives more inspiration to view neuroscience through the lens of new technology. Robinson has embraced virtual reality—now she's looking toward the next big tech innovation.

EyeWire is in the early stages of a generative sound design project with the Berklee College of Music. To the first-time player, its default Philip Glass-esque background music may seem numbing. It eventually synchronizes with the movements of the player's mouse, prodding her to keep mapping out the neuron. Now, the EyeWire-Berklee partnership will demonstrate how AI software can write its own music, with cues from the human players.

Related: Virtual Reality's First Person Shooter Problem

Scientists at Google and elsewhere are turning to the 30-year-old digital music standard MIDI to teach neural networks how to write music.

In May, Google research scientist Douglas Eck left his Silicon Valley office to spend a few days at Moogfest, a gathering for music, art, and technology enthusiasts deep in North Carolina's Smoky Mountains. Eck told the festival's music-savvy attendees about his team's new ideas about how to teach computers to help musicians write music—generate harmonies, create transitions in a song, and elaborate on a recurring theme. Someday, the machine could learn to write a song all on its own.

Eck hadn't come to the festival—which was inspired by the legendary creator of the Moog synthesizer and peopled with musicians and electronic music nerds—simply to introduce his team's challenging project. To "learn" how to create art and music, he and his colleagues need users to feed the machines tons of data, using MIDI, a format more often associated with dinky video game sounds than with complex machine learning.

Researchers have been experimenting with AI-generated music for years. Scientists at Sony's Computer Science Laboratory in France recently released what some have called the first AI-generated pop songs, composed from their in-house AI algorithms (although they were arranged by a human musician, who also wrote their lyrics). Their AI platform, FlowMachines, has also composed jazz and classical scores in the past using MIDI. Eck's talk at Moogfest was a prelude to a Google research program called Magenta, which aims to write code that can learn how to generate art, starting with music.

Listening to and making music is worth pursuing because, researchers say, both activities can help intelligent systems achieve the holy grail of intelligence: cognition. Just as computers are starting to evolve from simply reading text to understanding speech, computers might start to regularly interpret and generate their own music.

"You can learn an awful lot about language by studying text. MIDI gives us the musical equivalent. The more we understand about music creation and music perception, the more we'll understand general, important aspects of communication and cognition," says Eck, now a research scientist on Google's Magenta project.

From Crashing Computers To Making Them More Creative

As synthesizers gained popularity in the 1970s and 1980s, engineers started to experiment with ways to get their electronic instruments to communicate with each other. The result was the Musical Instrument Digital Interface, or MIDI, which the music industry adopted as a technical standard in 1983 after its creators, Dave Smith and Ikutaro Kakehashi, made it royalty-free, offering up the idea for the world to use.

"In hindsight, I think it was the right thing to do," Smith told Fortune in 2013. "We wanted to be sure we had 100% participation, so we decided not to charge any other companies that wanted to use it."

Personal computers soon evolved to read and store MIDI files, which reduce high-level, abstract pieces of music into machine-readable data in a very compact format (a song stored in a 4 MB MP3 file would be a mere few hundred kilobytes in MIDI). MIDI would become standard on electronic instruments, from keyboards and drum machines to MIDI guitar controllers and electronic drum kits. Music composed through MIDI has powered the rise of dance, techno, house, and drum and bass music, and its sound can be heard in most television and film scores.

MIDI is the symbolic representation of music, just like text is the symbolic representation of speech. "MIDI itself doesn't contain sound—it's just instructions," says Jonathan Lee, a musician who specializes in MIDI.

One MIDI link can carry up to 16 channels of data, indicating information about things like musical notation, pitch and velocity, volume, vibrato, audio panning, cues, and tempo. Instruments can also retrieve sounds from a set of prerecorded sounds, called SoundFonts, which are stored in a separate file. The format provides a wide latitude for musicians, allowing novices to compose sophisticated arrangements and more experienced talents to build complex orchestrations.

Though digital instruments still use the 30-year-old, five-pin MIDI connection, every modern computer and even the Chrome browser has the ability to receive instructions from MIDI devices via a USB adapter. The format has come a long way since 1990s Geocities pages and games like Doom, thanks to better computing power, digital samplers, and recent movements like "Black MIDI," in which MIDI musicians like Lee saturate a digital musical score with so many notes, typically in the thousands or millions, that little white peers through.

The most popular song that Lee has ever released—1.6 million views on his YouTube channel, TheSuperMarioBros2—contains 7.6 million musical notes. Playing it on YouTube sounds like Philip Glass on acid; playing it through MIDI software like Piano From Above or Synthesia would likely crash your computer.

"Ninety percent of computers wouldn't be able to play it without giving up," says Lee.

Lee, a 17-year-old Houston native, said he burned through the RAM and CPUs of both of his parents' laptops by experimenting with these Black MIDIs. He eventually bought himself a gaming-grade computer that could withstand his experimentation.

Lee believes that Black MIDIs, with their densely complex set of computer instructions, could push engineers to create software that relies less on RAM and more on CPU power, for example. This could help prevent computers from crashing during periods of heavy processing.

Deep Learning With Music

As learning material, MIDI files are a computer scientist's dream, unlike audio recordings; They are small, available in troves on the internet, and royalty-free, providing a resource that can be used to virtually train AI machines without limit.

The state of the art in training computers is deep learning, artificial learning that uses neural networks, a method of storing information that loosely approximates the information processing of the brain and nervous system. In computer vision, where deep learning has become the standard machine learning technique, scientists know how a computer learns through a neural network when the computer knows what shapes to look for in an image. You can see this process in reverse in the Deep Dream algorithm. Google engineers Alexander Mordvintsev, Christopher Olah, and Mike Tyka used the company's image-recognition software to "hallucinate" images out of everyday scenes, based on the system's memory of other images it had found online.

What perplexes scientists more is how and if computers can perceive something that is more subjective, like music genres, chords, and moods. Listening to music can help computers reach this higher-level cognitive step.

In July, a team of scientists from Queen Mary University of London reported they had trained a neural network to determine musical genres with 75% accuracy by feeding it with 6,600 songs in three genres: ballad, dance, and hip-hop. Then they tore apart the layers of the computer's neural network in order to see what the network learned at each layer when the scientists exposed it to songs from Bach and Eminem. The researchers saw that the computer started to detect basic patterns, like percussive instruments, at the neural network's lower levels, and more abstract concepts, like harmonic patterns, at the highest level.

Rather than using MIDI representations or other kinds of music notation, the researchers fed their learning algorithms 80,000 samples of raw audio signals extracted from the 8,000 songs. That decision may reflect the limitations of MIDI and other digital representations for teaching machines to learn the nuances of analog music.

In MIDI, the human voice is "that human intangible that ends up being a little bit absent, in the same way that you might have a Boston accent versus one from Texas or Minnesota."

"Something like MIDI has a lot of potential in terms of modeling classical elements of music such as harmony, or rhythm, or structure and form," says Eric Humphrey, a former doctoral researcher at the Music and Audio Research Laboratory at New York University, who is now a senior machine learning researcher at Spotify. "But what is really interesting is that MIDI isn't necessarily good at modeling timbre or production effects." Among other things, that means that "a lot of popular, modern music is not well-coded by MIDI."

"In terms of the expressivity and the nuance of the performance, that's that human intangible that ends up being a little bit absent, in the same way that you might have a Boston accent versus one from Texas or Minnesota," says Humphrey.

Rather than meditating on what might get lost in the art form, Google has already started to build new deep learning models to generate music. Over the summer, Project Magenta researcher Anna Huang designed a neural network to write in new voice sections in Bach chorales, whose original voice sections she had sporadically deleted. Huang and the team initially considered using the techniques of computer speech generation to finish the middle of a song if a musician had already written the beginning and the end.

But the researchers saw two problems with reusing the machine learning models used for speech generation. First, music is multifarious; several instruments and voices play out at once. In speech recognition, the computer essentially must only learn the pattern of one person talking. Second, musicians might not write scores in a linear fashion but, instead, opt to go back and fill in gaps as they compose. Spoken languages, on the other hand, build ideas in a logical sequence.

To address the first problem, the researchers took a cue from the field of image recognition. They found a machine learning model that taught computers to reconstruct the blanks in an image, a method called "inpainting." They thought that if computers could simultaneously recognize three RGB values in a picture, then they could think of each voice as a separate RGB value in their new model. To address the second problem, they decided to write an algorithm that let the computer generate melodies randomly, rather than sequentially.

The team trained its computers with several MIDI Bach chorales containing soprano, alto, mezzo-soprano, and baritone voice parts that they randomly cut out at different points in the pieces. At any given time during the modified sections, the computer would "hear" one to three voices. Then, the researchers tested what the computer learned by gradually taking out each voice part in a single Bach chorale until none were left. The team left the computer's 28-layer neural network to generate new voices from each previous voice it had generated.

In the end, Google researchers were pleased with the overall aesthetic of the computer's new work. Analyzing these Bach chorales taught Google that it could teach a computer to resolve musical dissonance, lean toward a preferred set of harmonies, and learn the musical key.

But their model only digitally approximated a few real-world musical style points. For one thing, their model did not account for the natural limits in pitch range that specialized vocalists, like a soprano or alto, would have. At certain points, the computer inflected the pitch in a voice line to stay in line with the musical key. The team is working on new ways to better code these human aspects into its machine learning model.

To do that, it will need more music to "teach" from. Besides producing new research useful for artificial intelligence more broadly, the Magenta engineers are also enthusiastic about growing their collaboration with the music community.

In August, the team issued an updated interface between musicians and TensorFlow, Google's open-source AI software. The new release allows musicians to connect Google's AI model to their own synthesizers and MIDI controllers to make AI-generated music in real time. Meanwhile software developers can also connect their own AI models in lieu of the one from Google, in the hopes that injecting non-Google ideas into the Magenta community will trigger more experimentation.

Lee, meanwhile, continues to make his own brand of Black MIDIs and publish them on YouTube. His MIDI compositions weave novel "note art," like swirls, letters, and even Morse code, into the visual effects of the musical score. Some are more mathematical in nature—one video, "Pi," contains exactly 3.141592 million musical notes and is 3:14 long, and another one, "Fractal Images," describes a set of mathematical equations called the Mandelbrot set.

When told that Google was looking for people to contribute MIDI files to its new AI project, Lee was eager to contribute. He plans to canvas the entire Black MIDI community to upload their files to the project. If those super-dense MIDI files don't crash computers, perhaps they could teach them a few things about how to write their own Black MIDI songs. "We're going to flood them with good content," he says.

This story was originally published with the headline, "The Music That Inspires Computers To Write Their Own Songs"

After a design theft in China and amid a growing market, how the tree-tent company Tentsile is making waves in camping and ecotourism.

Alex Shirley-Smith was a tree house architect. In 2013, fed up with designing custom-made suspended abodes that bordered on the extravagant for his wealthy clientele, he returned to earth and decided to find a new way up. With a friend, Kirk Kirchev, he set to work designing a tent that would perform much the same function as those houses: lifting people above nature and, he thinks, elevating the whole experience of camping.

Read Full Story

The company database CrunchBase has added gender to its profile pages, a precious bit of open data in the effort to diversify the industry.

On their own, the tech industry’s much-talked about “diversity numbers” don’t say everything about diversity in the industry, but they can indicate, for instance, in which cities women may be more likely to raise money (New York and San Francisco—more on that below). They can also offer insight into a phenomenon that’s mostly invisible. If you want to manage a problem, goes the engineer’s maxim, you have to measure it.

Read Full Story

Anti-piracy technology lets media industries trace movies, music, and books across the Internet. Can it help the digital art market?

Thanks in no small part to the Internet, digital art is having a moment, and it’s attracting collectors too. An auction last year of GIFs, digital paintings, and printouts at Phillips in London raised over $113,000, including $3,500 paid for a website by the Dutch-Brazilian Internet artist Rafael Rozendaal.

Read Full Story

The Seattle-based company has automated almost everything that goes into making its leggings (well, except for the sewing).

If programmer-inventor Pablos Holman had it his way, he would transform his clothing factory into one big 3-D printing machine. After all, he has tons of experience with 3-D printing technology, having helped create the machines at MakerBot. But with the technology being what it is, Holman will have to hold out until 3-D printers can better handle fabric.

Read Full Story