Russian cybersecurity researchers found that tech executives traveling across Asia may have had their computers infiltrated by password-sussing malware. No hotels or victims were specified. But the operation has been given the apropos spooky name "Darkhotel."
According to the Wall Street Journal, "Thousands of people traveling mostly in Japan, as well as Taiwan, China and other countries, have been affected by the attacks, which are likely targeted at a specific individual and occur when the traveler connects to the hotel wireless or cable internet." The attack works by asking users to install seemingly benign updates for software like Adobe Flash and Google's Toolbar. In actuality they are malicious trojans used to steal further information.
If you absolutely must use hotel Wi-Fi, Kaspersky Labs recommends using a Virtual Private Network, or VPN, to encrypt your communication channels, and to regard all software updates—no matter how safe they appear to be—as suspicious.
For something so centered around automation, programming is an awfully manual process. Every time a developer churns out a new line of code, there's a very good chance somebody has crafted the very same syntax—or something close to it—somewhere else in the world. Yet right this very moment, countless devs are toiling away at their own workstations, doing duplicate work and in the process, wasting precious time.
Now, the government agency that gave us the Internet is hoping to fix that problem. In concert with more than 20 computer scientists at three universities, DARPA is working on the programmer's version of autocomplete and autosuggest. With this advance, coding the next killer app will be as easy as Googling for GIFs or accidentally drunk-texting the wrong thing to your ex. Okay, maybe not that easy but you get the idea.
PLINY, which is named after the Roman encyclopedia author, will use Bayesian statistics to mine an enormous trove of open source code for common properties and behaviors. As a developer types, the system will mine this reservoir of code for logical clues about what might come next.
The system will also streamline the process of debugging and spotting potential security flaws, leaning on the collective intelligence of millions of programmers in real time rather than the fleshy slab of tissue in one person's skull.
The $11 million project—which is being led by Rice University—will unfold over the next four years as computer scientists figure out how to build a version of autocomplete smart enough to traverse billions of lines of code written in various languages.
...PLINY will need to be sophisticated enough to recognize and match similar patterns regardless of differences in programming languages and code specifications. The system will have to explore different ways of interweaving code retrieved through search into a programmer's partially completed draft program and analyze the resulting code to make sure that it does not have bugs or security flaws.
The core of the system will be a data-mining engine that continuously scans the massive repository of open-source code. The engine will leverage the latest techniques in deep program analyses and big-data analytics to populate and refine a database that can be queried whenever a programmer needs help finishing or debugging a piece of code.
Just as some code editors today automatically suggest specific tags or commonly used properties, PLINY aims to make much more granular and complex suggestions based on code that's already been written. If executed properly, it's easy to imagine such a system speeding up the development process pretty dramatically. Or, if our experience with autocorrect on our phones is any indication, it could introduce a whole new kind of headache.
The internet lost its collective mind this weekend over an 11-minute satirical sitcom intro clip called Too Many Cooks, which originally aired as segments on Adult Swim which is what Cartoon Network turns into at night, I guess? In a remarkable display of weekend tab-generating efficiency, we already have Behind the Sceneses, Making Ofs, Whatever, It's Not So Greats and What This Says About Yous. Buzzfeed's Jace Lacobturned in a Millennial-flavored thinkpiece that seems to start from the premise that you only saw 80s sitcoms "from the couch of your grandparents' house" and not, for example, when they were first-run primetime entertainment that you begged your parents to let you stay up late to watch (because you are old). Perhaps the strangest response, though, was Amazon announcing a Too Many Cooks tie-in home electronics product called Echo.
"He got 99 iPhones but gettin' hitched ain't done," writes Daniel Beanover at Yahoo Tech, in a story that's as funny as it is tragic. (Incidentally, that sentence is a strong contender for Lede Of The Week. Big shout!)
Anyway, here's the short version: In anticipation of Singles' Day—which somehow manages to be much happier than Singles Awareness Day, February 14th—a Chinese man bought 99 iPhones (worth a cool $82,000!), arranged them in a heart shape, and proposed to his girlfriend in front of a crowd. She said no. There are pictures.
The thirst cut is the deepest, I guess?
Lavish obnoxious public marriage proposals being shot down is literally my favorite thing, so thanks Bijan. You did good today.
Privacy has been pushed to the forefront of most people's minds in the last two years. A renewed awareness began in 2012 with the Snowden revelations, but its embers have been continuously kindled by non-governmental intrusions like the Target data breach and wariness over just what information apps like SnapChat and Whisper—as well as long-standing Internet giants like Facebook and Google—actually know about us and which of our personally identifiable data they decide to keep on their servers.
Perhaps it's no surprise then that new businesses have sprung up around privacy itself. But in the technology sector, monetizing privacy is hard.
Sure, there's startups like Blackphone and Vysk that make money by selling hardware, but where software is concerned it's difficult to turn a profit if you don't directly monetize your users' data (as Google, Yahoo, Facebook, and virtually every Internet giant does). Of course you can sell your services direct to users, but that's a hard pitch when you are trying to grow your user base; besides people expect software for free nowadays.
But one startup had a plan to promote, spread, and enable user privacy—and actually make a living from it. They created the Epic Privacy Browser, which allows you to browse the web in complete anonymity while at the same time showing them relevant ads from Google. Privacy and profit were hand in hand. That is, Epic's creator says, until Google changed its terms.
The easiest way to describe Epic is by saying it's similar to Tor, but extremely user friendly and works on the regular Internet. It's similar to Tor because it allows anyone to browse online while at the same time allowing them to remain completely anonymous to the sites they visit and the services they use. But unlike Tor, which allows you to browse the dark web, Epic is an everyman browser that enables you to surf the web you already know and love. It's also dead simple to install and use with virtually zero setup—just download and go.
Like most things on the web, Epic has indirect ties to Google. First, it's based on the Chromium source code released by Google, which is also what Google Chrome is built on. Second, Epic uses Google ads, which are displayed to users when they search. This is how Epic makes money to pay its developers to continue working on Epic to enable a more private online experience for everyone.
Chrome also has a privacy mode built into the browser—it's called Incognito. However, contrary to what many might assume about Incognito, the privacy mode in Chrome does not make you anonymous to most websites or Google itself. Incognito does not prevent against most remote tracking of search histories or other information about you obtained via cookies. What Incognito does do is protect your Internet usage history from others who may be using the same computer you are on. It's local protection, not an online anonymity tool.
As for ads, we all know Google is an ad company and you, the user, are its consumer. Google's services like Maps and Search are both a benefit to the end user and also a way for the company to record more information about their product (i.e., you) so they can package and sell more relevant info to their customers, the advertisers. Sure, it sounds kind of creepy when put that way, but it's a valid business model and it has made Google one of the most profitable and powerful companies the world has ever seen.
Epic, on the other hand, is the exact opposite of Google and Chrome. Its browser destroys all cookies, history, and cache every time you exit the browser, while also maintaining complete anonymity by running all your searches and web usage through a proxy. The result is no sites or advertisers can track you or personally identify any characteristics about you. For what it's worth Firefox has just announced a "forget" button that mimics Epic's features, but doesn't have the advantage of running your browsing history through a proxy.
As for the ads Epic serves to its users, they are designed to still be relevant while not compromising a user's privacy because Epic masks every single user's IP address. Epic then provides the masked IPs to one of Google's ads providers, a company called Blucora that sells private-label Internet search services—including Google ads—to companies. Blucora then returns ads relevant to the Epic user's geographic location without Epic, Blucora, Google, or the advertiser ever knowing where specifically a user is—or where they have browsed online.
It is perhaps an inelegant solution, but one that solved the problem of keeping a user anonymous while still providing value for advertisers and a revenue source for Epic.
"We had a great deal where we would send a masked IP address to them to get the ads. That was that system that we had set up with them," says Epic's CEO and founder Alok Bhardwaj, 39, when I ask about the original arrangement that Bhardwaj saw as a win-win for both companies. Bhardwaj is no stranger to negotiating deals that benefit all parties involved. Before Epic he started out in finance advising technology companies on mergers and acquisitions as an investment banker with a boutique acquired by the French bank Societe Generale. After leaving that he worked as a teacher and tutor and also authored research papers on the telecommunications, technology, and gaming industries at the Columbia Business School. "Google needs to know the IP address of whoever the person is that wants to see the ad, so they know if it's relevant for the advertiser and if that's what this advertiser is really paying for. So that's why Google wants the IP address. But then when we masked it, but we would mask it in such a way that Google still gets all this geo location data; they know where this person is, but they just can't personally identify that person."
And things were going great for almost a year until one day Blucora informed Epic that Google was now requiring the true IPs of every user who viewed their ads. Epic refused and Google cut off their ads. And like that, the privacy browser's revenue sources were severed.
When Blucora informed Epic that Google was cutting off its ads Bhardwaj says he tried to reach out to Google, but had no success. He also says he worked with the ad intermediary Blucora to try to convince Google to go back to its original agreement.
"Our partners were trying to help us," says Bhardwaj. "They were basically trying to talk to Google and say, 'Hey, you know it would be great if you guys could continue to support the same kind of masked IP address privacy-protecting measures that you had supported before.' They were trying to get Google to support all of that, but Google basically just said 'no.'"
Blucora declined an invitation to be interviewed when contacted, only asking if I'd spoken to Google.
A Google spokesperson did confirm the changes in its terms now requiring actual IP addresses for ads, but said that these changes were not directed at any one company. Instead the changes to the IP requirements were just one part of a set of changes to improve the ad experience for both advertisers and users. Google says the changes made it faster to load ads on their sites as well as making it easier to push updates to partners very quickly so everyone has access to the same quality of ads.
As for changing the terms of requiring real IPs instead of masked IPs, the Google spokesperson said that was done because verifying actual IPs is a very important signal the company uses to weed out fraud—something that costs online advertisers $6 million a month. The Google spokesperson was also quick to point out that none of its changes were directed specifically at Epic nor are they suggesting in any way that Epic was involved in any kind of fraud. They also noted that Epic was not in violation of any of Google's policies, it was just the technical implementation of what Google needed on Google's end to make the ad systems work.
Bhardwaj says Google's explanation is "not very compelling" given that there are numerous methods to detect fraud. Yet while he knows exactly what he needs to do for Epic to start making money again—hand over Epic's user's real IP addresses to Google—it's not something he's willing to do.
Instead, Bhardwaj has filed an official complaint with the Federal Trade Commission, saying the company would like to see regulatory oversight requiring Google to provide search ads to all competing browsers if the browsers are willing to pay for those ads.
Nothing to hide: Alok Bhardwaj at TEDxChristUniversity
In spite of the monetization problems Epic has encountered in trying to make privacy a profitable business model, Bhardwaj says he is not deterred. He's currently collecting donations to cover development costs as well as speaking with other ad providers.
"We do continue to talk to Microsoft and Yahoo, but both of those guys as well are actually not particularly inclined to support private search ads for us," he says. "Our hope is that as we keep growing, that maybe we can motivate those guys from a financial perspective to move, but neither of them are particularly inclined either to support private search ads."
There are other businesses that could benefit from a privacy browser that also serves ads, such as Snapchat or Ello. But those companies are still figuring out their own revenue models, and aren't in a great position to champion the Epic cause.
Nevertheless, Bhardwaj believes business models that support consumer privacy are inevitable, and Epic will be there to support them.
"Privacy is essential to freedom. We do thousands of searches and visit tens of thousands of websites a year. We send to the Internet now our stream of thought. Google knows more about you than your mother or your spouse if you use Gmail, Chrome, or Google search," says Bhardwaj. "We believe we're not the only ones uncomfortable with that and believe there should be consumer choice."
After six weeks of rigorous 9-to-5, Monday-through-Friday class time, HappyFunCorp's six-week long HFC Academy has taken a group with negligible coding skills and turned it into a squad of wee entrepreneurs-in-training who know their way around a website product. Now we ask the really important question: Was the class worth it?
All coding bootcamps promise some level of proficiency by the end of the course, but what sets HFC Academy apart is its combination of coding lessons and product experience—especially when coupled with our final projects. The instructors pointed out time-sinks and places to strategically cut corners so as to sacrifice a minimum of polish in order to deliver complete products on deadline.
I entered knowing just the straggles of HTML from a 7th-grade class and left HFC Academy with the skills to build a website, and some product know-how to help navigate client negotiations and internal team management. Spending weeks struggling to get projects to function up to spec, I gained new respect for the need to protect engineers from the client's whims. The tough love I got from the instructors about being too precious with code has prepared me to stare down my hypothetical engineer and tell them that the feature they worked on for three weeks is kaput.
Since this is the only bootcamp I've completed, it's impossible for me to compare it to other programs. I believe the same is true for my fellow students, and in casual conversation they told me they were very satisfied with HFC Academy. They loved the product lessons and exited with some coding skill to build simple versions of the complex products HappyFunCorp builds for clients.
As tech entrepreneur Karl Hughes points out, bootcamps have become a "get successful quick" solution for folks who want to shift careers and perhaps throw their hat in the startup ring after just a couple months of intense instruction. One of the things that Hughes hasn't seen these bootcamps doing is being transparent with how long it takes to become a good web developer. To HFC Academy's credit, we were never given the delusion that we'd exit with the skills to walk right into a coding job.
Instead, the class did exactly what Hughes says bootcamps do well: introduced us to the product development world, introduced us to the startup and web development industry, and gave us entry to HappyFunCorp's network. HFC Academy also delivered on exposing us to a wide range of product lessons, from professional product expectations and standards to bringing in guest speakers to speak about programming and startup culture. The speaker lineup was less robust than the continuous lessons taught by the HFC Academy instructors and HappyFunCorp developers (one guest speaker couldn't come and wasn't able to reschedule), but the course taught me lessons I wouldn't be able to find elsewhere—and likely not in the more business-agnostic coding bootcamps out there.
Bootcamps are somewhat new and difficult to compare, so the de facto rule with them is unfortunately still caveat emptor. Hughes emphasizes that all bootcamps are different, so do your research to see if one is better for you than another. Want a six-week intensive Ruby on Rails course honing product strategy? HFC Academy might be right for you—but you might want to look at what else HappyFunCorp has cooking before you carve six weeks out of your life.
Parting Lessons
Our last guest speaker was HappyFunCorp cofounder Ben Schippers, who'd run the proto-HFC Academy at Bates College and developed the first HFC Academy's curriculum alongside HappyFunCorp's other cofounder, Will Schenk.
After weeks of guests and lessons about project and startup management, Schippers focused on the biggest obstacle standing in the way of success: ourselves. Over the years building startups and products for myriad businesses, Schippers keyed in on how good entrepreneurs take care of themselves. Some of this was quite simple—do what you love!—and like all simple things, very hard to stay focused on.
"One thing that's very hard is figuring out how to move the needle for yourself, not for other people," Schippers says. "Even if you don't want to start your own business, this is where you tend to fall apart. This is where people 'succeeding' tend to do well."
To maintain focused work on your personal goals, make an incredibly simple to-do list.
"When I'm starting a company, for the first two years, I'll wake up and write down three things to do. I don't put a fourth thing—don't even kid yourself," Schippers says. "Try to focus on three things that are really going to move the needle forward for you. You'd be surprised how hard it is to do."
Those shortlists, sleep, and time management are critical to maintaining focus, Schippers says. Make them priorities. Exercise and nutrition are also important. These are obvious cornerstones for personal health, but they're also the corners people tend to cut quickest when the hours start ramping up.
"Get comfortable with knowing that you're not gonna know what's gonna happen tomorrow. If not, you're never gonna push yourself to get off," Schippers says. "It's accepting the fact that you're gonna have a lot of unknowns, which affects focus. Which is why sleep and time management and goals are important."
Today's tech landscape is different than the first dotcom boom, Schippers says. Things used to be much harder to build, but now you can make products with a handful of people on a team. The downside is that everyone has made those products and the market is saturated. There's no guarantee anybody will use what you make. So unless you've got a technology-revolutionizing app, you need to differentiate your idea, or get lost in the sea of app stores.
"For those of you who think if you build a technology people will use it, they won't. You need to be working on people," Schippers says. "It's not a technology problem, it's a distribution problem."
To build momentum, to stride in a saturated sea, you need to fail fast, Schippers says. If you spend $100,000 on a product that sucks, he says, move on. The great news is that you can make that $100,000 again. But you can't get that time back, so iterate quickly and get feedback quickly. You need choice reviewers who will keep you honest and not just lavish you with platitudes.
"Do yourself a favor: Make sure you're talking to people about your idea and make sure they remove one feature," Schippers says.
You don't just need to pump your network for feedback, you need to pump your network. Go to events, connect on LinkedIn, go out for coffee. If somebody's willing to have a meeting, take it, no matter who they are, Schippers says. Just make sure that person is going to make an introduction into their network. That's how you create opportunities.
While you're sprinting forward, follow the old adage of asking forgiveness instead of permission.
"Don't ask permission, just move faster. As you begin to play chess with the world around you, you're gonna piss people off. Get over it. If you really piss people off, apologize. I push that really far," Schippers says.
It'll take you time to build and iterate and sell your product. It might take four months to build, but it'll take 15 months to get the product right. Remember that Instagram, Twitter, and Airbnb weren't built overnight, Schippers says. Some are, but don't try to be them: You'll just feel like a failure. As an entrepreneur, it all comes down to grit and managing discomfort. This is where a lot of people fail, Schippers says.
"There's interesting literature about people who are very successful: They don't talk about business—they talk about managing the lows," he says.
Exercise and social media vacations are Schippers' bread and butter. Get away from the strain of communication devices (Schippers even considers email a social tool, and won't answer past 7 p.m.). Pay attention to what you're putting in your body: people drink coffee, eat a lot of simple carbs, and then hit spikes of energy and crash. Better to find an even keel. You will become more efficient, Schippers assures. That balance extends to work: Find hobbies. Take walks.
And at the end of the day, you have to love what you're working on. Schippers has never seen it end well otherwise. Find meaning. Why are you doing what you're doing? Are you doing it for yourself or for someone else? Make sure you're making this product because you can't imagine doing anything else.
"I think money is a great reason. If money is what gets you up in the morning, guess what, it's a motivator. Not everybody is gonna save puppies and kittens," Schippers says. "A lotta people fail because they're constantly off-track. If this is what you want to do, this is what you should do."
Next Steps: For Students And The Academy
I asked the instructors what the value of a six-week product coding class should be.
"I would say it provides an accelerated way to understand problems you'll face when building something," program director Aaron Brocken says. "It helps build a skillset beyond being able to program, a real-world exposure to all the cuts and bruises you wouldn't get coding online at home."
Those bruises and cuts usually scare people off the coding game, especially if they're learning at home without experienced programmers pushing them in a learning environment.
Sadly, learning to code isn't a magic wand. Most of it will still be forensic debugging of things that should've worked, but didn't.
"The vast amount of 'coding' is debugging—trying to find a way to think around this thing," HappyFunCorp cofounder Schenk says. "You do know the shape of things more as you get better, but you always start debugging the same way."
The ultimate question for HappyFunCorp: Is this class worth putting on? The interest they've gotten since promoting the class proves that it's filling a need. But instead of expanding their six-week immersion courses, they're responding to demand at times and lengths working folks can access.
Six weeks is a long time to pull out of your life, Schenk says. So instead of filling the year with more immersion courses, they'll be hosting more event and two or three week seminar-style evening courses to meet interest.
"We've been getting a lot of inquiries: 'We love what you're doing, but can you do it at night?'" Schenk says.
That lets HappyFunCorp serve a much larger student body while the shorter seminar length will let them dive into more topics at different skill levels. But that doesn't mean the six-week immersion class is going away. On the contrary, it's still there to possibly answer the million-dollar question: Can they hire someone right out of the class?
They haven't—yet—though they're keeping an eye on the first class's students and advising them on post-class projects. The bottom line: The students need more product experience.
"The reason people are not going to get out of this course and immediately get a job at Facebook is that they just don't have enough product experience," says HappyFunCorp cofounder Schippers.
Hence HappyFunCorp's new focus on seminars and events that teach product lessons HFC learned over years of building websites and apps for clients. The shorter courses will also be attractive to prospective students who don't necessarily need to learn coding—they may just want to understand it to be more effective in their workplace, Schippers says. It'll provide exposure for people who don't necessarily need to dramatically change their careers.
"We want product stuff that conforms to resources they have," says Brocken. "So the after work bits and bobs events and after work seminars not only feeds people of different skill levels, it helps them in a safer environment by keeping their jobs."
There's a small chance if you took your entire music collection and mixed it all together that it might sound decent, but more likely it'd sound like a trash compactor. Yet, somehow Gregg Michael Gillis, better known as Girl Talk, has built a career on mashing up several hundred sampled songs into something that sounds like a cohesive thought.
Once you know you're listening to parts of different songs—that you may or may not know—the thoughts begin racing, "Wait, I recognize that sound, who is it?" And whether you realize it or not, it's the type of music that demands a visual layer. Matt Adereth has now added that visual element with his newest creation, an annotated guide to Girl Talk's last album, All Day.
Using illegal-tracklist.net (a completely legal site FYI), Adereth wrote a parser to effectively drive the visualization. Almost all samples from the album are included, though it isn't 100%. Doing the album annotation as a side project, it took hum a few weeks to complete.
"The biggest challenge was getting the years for the tracks that don't have Wikipedia pages, which required me to do some manual searching," Adereth says. "Eventually I found Who Sampled, which had all the details I was missing."
The result is a timeline filled with little blue data points, each representing a different sample. As the song plays and passes over each point, the sampled song's album cover pops up below and then disappears when it's done. The albums are then stacked by year they were released—which allows you to see which samples are driving the songs and feel.
For the project, the music is played using SoundCloud's Widget API, while the art is grabbed from Wikipedia and links via iTunes. On the more technical side, Adereth used d3.js for the visualizations and Clojure for the data processing.
If you want more, Adereth also has a version one concept for the Girl Talk album Feed The Animals.
Shingy is an older Australian gentleman huddled beneath the scalp of a sickly emo lion, whose job is to spout bullshit to terrified #brand executives who know they contribute nothing of value to the world and are merely riding a diminishing pile of money into oblivion, like starving polar bears drifting south on a melting ice floe. Shingy is extremely good at his job–he has plugged his proboscis into AOL's cash-hole to the widely-reported tune of "six figures" a year–and the last couple days have seen a bit of a Shingaissance, with delightful profiles by Andrew Marantzon the New Yorker dot com and Maria Bustillosin Details. While Shingy himself is a big, soft, stupid-looking target for ridicule, he's just a man who found his grift and is pumping everything he can out of it. If you think Shingy is ridiculous (and he is) why not cast an eye on the man who pays his salary, AOL CEO and notorious baby-haterTim Armstrong. "I love the smell of this table!" Armstrong exclaims in the New Yorker. I submit that Shingy knows exactly what he's doing, and it's Armstrong who is the hapless, furniture-huffing idiot.
Scocca also links former New Republic editor Sullivan to the New Republic's current 100th birthday celebration, formally titled: "The New Republic Looks Back On a Century of Undermining the Left From Within." Hanna Rosin is looking back on journalism's most spectacular liar and her former friend Stephen Glass in a profile / apologia / reminiscence–a profologiscence, if you will (rhymes with "excrescence"). I get the feeling Rosin is the only writer on earth who wouldn't have just filed "Of course Stephen Glass isn't entitled to become a lawyer, lol" on a Post-It note for this assignment. Also in famous liars: Plagiarist and apology-profiteerJonah Lehrer is working on a book. Please try to get your vom into the bucket, thanks.
There are some good things. There are many more bad ones. Here is a very bad thing that I have brought you, not unlike a cat that gifts you a dead animal on your doorstep, or a dog that has eaten some shit and vomits it onto your shoes as you're cleaning up what the cat's left for you.
This week's issue of the newly-paywalled New Yorker has Paige Williamsgoing very deep into an odd—and frankly unbelievable—protrusion of judicial power: judicial override. Judicial override is legal in exactly three states, Delaware, Florida, and Alabama; judicial override is exactly what it sounds like. Which is to say that judges—who, in blood-red Alabama, are democratically elected and are therefore subject to the esoteric moral whims of an insular group of white folk—can unilaterally disregard a jury's decision in capital cases.
Generally, as in less-crimson Florida and downright-blue Delaware, this is used to convert death sentences into life in prison without the possibility of parole. In Alabama, however, judges appear to use their demigod powers to murder people who are disproportionately black and poor.
Williams wends her way through AL and tracks the case of one Shonelle Jackson, who will probably be the first person in American history to be executed despite a jury's unanimous vote for his life. It is sickening. The quotes Williams has gotten, on the record, are sickening. There is nothing good about this, and it is hard and necessary reading. Ban the death penalty. Ban state-sponsored murder.
Deep learning is one of the buzziest topics in technology at the moment, and for good reason: This subset of machine learning can unearth all kinds of useful new insights in data and teach computers to do things like understand human speech and see things.
It employs the use of artificial neural networks to teach computers things like speech recognition, computer vision, and natural language processing. In the last few years, deep learning has helped forge advances in areas like object perception and machine translation—research topics that have long proven difficult for AI researchers to crack.
Trouble is, deep learning takes a ton of computational power, so its use is limited to companies that have the resources to throw at it. But what if you could achieve this kind of heavy duty artificial intelligence in the browser?
That's exactly the aim of a project out of Standford called ConvNetJS. It's a JavaScript framework that brings deep learning models to the browser without the need for all that computing muscle. In time, it could make your tabs a lot smarter.
"Deep Learning has relatively recently achieved quite a few really nice results on big, important datasets," says Andrej Karpathy, the Standard PhD student behind ConvNetJS. "However, the majority of the available libraries and code base are currently geared primarily towards efficiency."
Caffe, a popular convolutional neural network framework used by Flickr for image recognition (among many others) is written in C++ and is slow to compile. "ConvNetJS is designed with different trade-offs in mind," says Karpathy.
So how does this actually play out in the real world? Right now, Karpathy's website points to a couple of basic, live demos: classifying digits and other data using a neural network and using deep learning to dynamically "paint" an image based on any photo you upload. They are admittedly geeky—and not especially practical—examples, but what's important is the computation that's happening on the front end and how that's likely to evolve in the future.
One likely usage is the development of browser extensions that run neural networks directly on websites. This could allow for more easily implemented image recognition or tools that can quickly parse and summarize long articles and perform sentiment analysis on their text. As the client-side technology evolves, the list of possibilities for machine learning in the browser will only grow.
Because it's JavaScript-based, the framework can't pull off quite the computational heavy lifting that other tools can, but it nonetheless raises the interesting prospect of bringing machine learning directly into the browser. "The idea is that a website could train a network on their end and then distribute the weights to the client, so the compute all happens on client and not server side, perhaps significantly improving latencies and significantly simplifying the necessary codebase," Karpathy explains.
Tech companies have been getting more interested in deep learning only recently. A foreign topic to most, deep learning started inching its way toward mainstream buzzword status earlier this year when Google bought a deep learning startup called DeepMind for $400 million. The learning algorithms of DeepMind fit perfectly into Google Brain, the company's broader quest to emulate the human mind using machines, something Facebook is known to be working toward as well.
For years to come, the truly groundbreaking work in this subfield of AI will continue to happen on the back end and where the computational power is most potent. But as the field evolves and browsers themselves get more powerful, we can safely expect to see more of this sophisticated machine intelligence shift to the front end and, in all likelihood, supercharge what happens in our ever-expanding sea of browser tabs.
"ConvNetJS is not where I want it to be," admits Karpathy. "I work on it on a side of my PhD studies and with many deadlines it's hard to steal time. But I am slowly working on cleaner API, more docs, and WebGL support. Regardless, I think the cool trend here more generally is the possibility of running (or training) neural nets in the browser. Making our tabs smarter."
Indie games are often noted for their simplistic graphics, but Plotkin's game goes one step further: It has no graphical interface at all. Players explore a crashed starship, solve puzzles, and guide their characters through complex magical rituals using an interface based solely around the written word.
Hadean Lands is one of the latest examples of a nearly 40-year-old style of video game known as "interactive fiction," where players read novel-like descriptions of in-game action and either type their characters' actions or choose them from a written menu.
Until recently, interactive fiction seemed as though it had peaked in the 1980s, with new titles mostly made and played by a small but loyal group of fans. But in the last few years, new distribution channels and incredibly user-friendly development tools have not only revived the medium but brought in new players and developers who've pushed the genre far beyond its swords-and-sorcery roots.
"There's a greater variety of not just types of game content, although there's that too, but there's a greater diversity in terms of creators," says Jason McIntosh, a longtime player and creator of interactive fiction and the 2014 organizer of the annual Interactive Fiction Competition, now in its 20th year.
"When we're talking about IF, we're talking about games that are built entirely from text," says McIntosh, using a common acronym for interactive fiction. "Text—more so than making 3-D models and figuring out pathfinding and collision detection and particle effects and such—text itself is the one of the most accessible ways that anybody can create."
Fans trace the history of interactive fiction back to Colossal Cave Adventure, a game created in the 1970s by programmer, spelunker, and divorced dad William Crowther, partly as a way to bond with his young daughters. Crowther shared the game with friends and colleagues, and various clones and derivatives soon spread across the fledgling computer networks of the time.
Some of the MIT-based creators of Zork, one of the more polished games inspired by Crowther's classic adventure, went on in the late '70s to found a software company called Infocom, making interactive fiction games for early home computers.
By its mid-'80s heyday, Infocom's titles ranged from an adaptation of A Hitchhiker's Guide to the Galaxy, co-created by the book's author Douglas Adams, to The Witness, a 1930s Los Angeles noir-inspired detective game, to A Mind Forever Voyaging, a dystopian time-travel adventure.
"For the first time, you're more than a passive reader," said marketing materials for A Mind Forever Voyaging. "You can talk to the story, typing in full English sentences. And the story talks right back, communicating entirely in vividly descriptive prose."
Selling cross-platform games in the era before Microsoft cornered the market on consumer operating systems, Infocom built its games to run on a virtual machine whose code could be interpreted on a variety of computers, just as Java would later enable programmers to write applications that ran on Windows, Mac, and Unix systems.
But by the end of the '80s, Infocom and other interactive fiction publishers struggled to compete with increasingly sophisticated graphical video games. Adventure games featuring puzzles and exploration stayed popular—players snapped up installments from Sierra On-Line's King's Quest franchise, LucasArts' madcap adventures like Monkey Island and Maniac Mansion, and CD-ROM-powered, graphically rich offerings like Myst, the best-selling PC game of the 1990s—but those games largely replaced text with animation, and typing full English sentences with pointing and clicking the mouse.
Still, fans remained committed to text adventures, and in 1993, developer Graham Nelson released Inform, an open source programming language designed for crafting interactive fiction and, crucially, capable of generating code that could run on Infocom's virtual machine.
"Infocom game story files are as near to a universal format as we have for interactive fiction games, but until now it has been very difficult to construct them, and I am not aware that anyone has previously created them outside of Infocom itself," he wrote in a post to Usenet, the premier pre-web network of Internet forums, announcing Inform.
And in 1995, members of the Usenet interactive fiction community launched what would become the annual Interactive Fiction Competition. They wanted both new games to play and a way for hobbyist developers to swap examples of quality code in Inform and a similar-purpose language called TADS.
"I LOVE IF and am glad there are people just like me!" said one post in response to the contest proposal. "Thank you, everyone, for keeping an underappreciated hobby alive!"
Plotkin, who'd launch Hadean Lands nearly 20 years later, won his division, and he, Nelson, and other developers continued work on better tools to power more sophisticated interactive fiction.
"Basically, the current movement has its roots in the mid-'90s with both the [competition] and Graham Nelson," says McIntosh.
The genre continued to draw in players and creators who grew up playing Infocom's games in the '80s, then younger audiences who grew up on LucasArts graphical adventures and open source interactive fiction, he says. But game authoring systems still required authors know some programming, and the games' commands parsers, usually expecting directives in a distinctively terse dialect of English, could be a barrier for new players.
"The reason the parser works the way it does is it came out of 1970s computer culture where, how did you interact with a computer? You typed things into the command line," he says. "Unless you're like a system administrator [nowadays], you're not used to typing in commands on a command line."
But in 2009, developer Chris Klimas released a tool called Twine, which made it possible for non-programmers to create interactive fiction by drawing flowcharts showing game scenes and the player choices that propel the game in different possible directions.
"It only started blowing up just in the last two years," says McIntosh. "Twine is definitely the [area] where there is the most community attention, and there is the most innovation happening right now."
Twine compiles its flowcharts to HTML, with transitions represented as ordinary web links, meaning IF newbies no longer had to worry about downloading virtual machines and compilers. Twine games are simply web pages, able to be uploaded to any old hosting site, or Twine-specific repositories, and played in an ordinary browser.
More sophisticated Twine games can be made with custom HTML and JavaScript, but it's certainly not necessary, and the tool attracted new creators and players often interested in shorter, more personal games of just a few scenes, often capturing creators' thoughts and emotions.
"This isn't just about about different game *interfaces*," Plotkin said in an email. "These are different communities, different audiences, different artistic conventions."
One game listed on the Twine site, called You Are Invited, sympathetically portrays a character's struggles with anxiety. Another, called Violet, drew critical acclaim for its story of a 30-year-old graduate student struggling to finish his dissertation. The story's narrated in the second person by the student's exasperated girlfriend, determined to leave him if he can't get his act together.
"[Twine] games tend to be intensely personal and direct," wrote Plotkin. "Not *all* of them — but the 'archetypical' Twine game is short, expressive, meant to convey one experience directly from the author to the player."
The tool has let people who would have never previously thought to make video games, or even to search for video games with characters they can relate to, enter the field and express themselves, says game developer and writer Anna Anthropy.
"I've played hundreds of Twine games by people who never made games before, maybe didn't play games before," says Anthropy. Her book, Rise of the Videogame Zinesters, discusses how Twine and other new authoring tools let would-be game creators without programming training or a traditional game-developer background create games, just as zine culture created an outlet for lo-fi, homemade print publications.
"There are a lot of people who I would like to see making games, like a lot of marginalized people who I think would use the form and are using the form in amazing ways," says Anthropy. "I think that games, or interactive work in general, lends really well to telling stories that are about systems or struggling with or engaging with those systems."
Anthropy's own games range from Star Court, a piece of Choose Your Own Adventure-type interactive fiction with a style of humor reminiscent of Douglas Adams, to Dys4ia, based on Anthropy's own experiences as a trans woman, to Encyclopedia Fuckme and the Case of the Missing Entree.
In the latter game, the player guides a female character, loosely inspired by Donald Sobol's classic boy detective Encyclopedia Brown, as she tries to escape from her lover, who she discovers is a cannibalistic serial killer. Using text lets Anthropy explore the character's complex and changing blend of attraction and fear as she uses a mixture of seduction, deception, and brute force to make her escape.
Another of Anthropy's interactive fiction games, Queers in Love at the End of the World, portrays a character with just 10 seconds to say goodbye to a beloved partner. It deliberately bombards players with an overwhelming array of choices, thoughts, and sensations as a timer ticks away Earth's last seconds—a sensation she says it would be hard to express through another medium.
"You have exactly 10 seconds and an enormous swath of different options—of all branching, different options, that are completely unnavigable in 10 seconds," she says. "I don't know how, that sense of urgency, how I would be able to create that in a more static form."
Some of Anthropy's games are available online for free, which is something she says is made economically viable by her supporters on the crowdfunding site Patreon.
"Patreon has been a big help to me," she says of the site, which emphasizes letting backers make steady contributions to artists regularly producing work they enjoy. "At this point, Patreon basically pays my rent."
Others are sold through indie game markets like Itch.io, where she says some fans are willing to pay more than the base price to support game makers they like.
Online markets, from indie sites to the big mobile app stores, are a boon to independent developers, as are more flexible crowdfunding models like Patreon's, says Plotkin.
"In 2010, Apple's App Store was an incredible deal — a way for a solo developer to offer a game worldwide, no contract negotiation, all payments and distribution taken care of," he wrote in an email. "Today I can do the same on desktop OSes using the Humble and Itch.IO platforms."
Plotkin's built open source tools to turn traditional virtual-machine IF games into iOS apps, and, he says, the mobile versions of Hadean Lands include features like a tappable directory of frequently used commands, to minimize typing, and a built-in virtual scratch pad.
And just as tools evolve to take advantage of new technologies, games will continue to evolve as long as their creators continue to be willing to experiment, says game developer Mark Musante.
"There's still plenty of things to try, even though we've been at this for around 40 years," he wrote in an email. "Compared to, say, writing books, this style of storytelling is still in its infancy."
If you're reading this, chances are you have a graveyard of defunct gadgets at home. Phones. Old audio equipment. What are you going to do with all that junk? You could always ship it off to get recycled, but that's boring.
Why not make something new out of it?
Like a whole genre of music.
That's what musician and sound artist Yuri Suzuki and his collaborators decided to do after scouring the streets of Johannesburg, South Africa and uncovering a trove of old gadgets. Using the old tape recorders, cell phones, and other electronics they found, Suzuki and a team of artists hacked together playable musical instruments and recorded their own version of the early minimalist electronic song "Warm Leatherette."
Suzuki teamed up with fellow artists Bogosi Sekhukhuni, Nathan Gates, and Neo Mahlasela to create the instruments and produce the song and accompanying video, which can be seen below.
"I am curious about the music scenes of each country and South Africa has very interesting music scenes such as Shangaan electro," says Suzuki. "I thought would be nice to invent a genre during this commission. So we came up the idea to make genre from scratch, specifically a genre which is made from DIY musical instruments."
This isn't the first time Suzuki has tinkered with conventional notions of what musical instruments are and how they work. Earlier this year, Suzuki and collaborator Mark McKeague started shipping the Ototo, a DIY synthesizer that uses sensors to turn everyday objects into touch-controlled musical instruments. The device allows people to create instruments from just about anything, whether by hooking it up to pieces of fruit or using it as a MIDI controller to achieve essentially unlimited sonic possibilities.
This latest project similarly takes common objects and turns them into tools for creating music. It was commissioned by digital creative agency onedotzero as part of a partnership between the Department of Arts & Culture of South Africa and the British Council.
Yesterday, a package of 10 "beautiful yellow round pills with the Twitter logo" arrived safely at an art gallery in Switzerland. The pills were sent from somewhere in Germany—the actual origin is unknown—and were vacuum-sealed in aluminum foil and hidden inside a DVD case. The "snapback 120mg MDMA," as the pills are formally known, was purchased by the "Random Darknet Shopper", a bot created by the !Mediengruppe Bitnik art collective for "The Darknet – From Memes to Onionland. An Exploration" exhibition at Kunst Halle St. Gallen gallery in St. Gallen, Switzerland. The exhibit offers a fascinating glimpse of the World Wide Web's underside.
"Random Darknet Shopper" is an automated bot, an electronic shopper that gets a weekly budget of $100 in bitcoin. Among its recent purchases (all of which are on display at the gallery): spy gear, a Lord of the Rings e-books box set, 10-packs of Chesterfield Blue cigarettes shipped from the Ukraine through Moldova to St. Gallen, a Platinum Visa card and a pair of Nike Air Yeezy2 sneakers. So far, !Mediengruppe Bitnik's favorite item is the Firebrigade Master Key Set from the U.K. "It has this mystical quality as to what you can do with it," say Domagoj Smoljo and Carmen Weisskopf of !Mediengruppe Bitnik. (Potentially, the master keys could open locked subway systems, allow access to firehouses, and even operate old-time elevators.)
Most of these items aren't the kind of things you can pick up just anywhere online. Agora and other "deep web" markets can be accessed via Tor, technology meant to protect users' anonymity, and cloak their activity from snooping government bodies. The artists describe Random Darknet Shopper as a live mail art piece, similar to their "deliveryseries" which began with Delivery For Mr. Assange when they shipped a parcel to the notorious WikiLeaks editor at the Ecuadorian embassy in London. That package had a tiny hole in it, through which a hidden camera documented its trip through the postal system. This current project has a safe and legal predecessor: Darius Kazemi's Random Shopper bot, which randomly purchased an assortment of stuff from Amazon. But things get ethically and legally murkier—and quite a bit more interesting— when a crawler specifically devoted to Agora can trawl through its 16,000 items, many of which are as illicit as Ecstasy.
"We want to see what goods come out of the deep web," say the artists, "where they are sent from, how (and whether) they arrive. We want to find out how the goods are packaged to be concealed from the postal services."
The Random Darknet Shopper raises all kinds of conundrums. How can a society manage "dark" markets given the problems that arise when an international online marketplace runs into the different laws of different countries? Then there's new relatively new phenomena of these "dark" markets adopting capitalist business practices, and having to rely on trust between sellers and buyers. "Although the hidden market is based on the anonymity of its participants, rating systems and anonymous message boards ensure a certain level of trust," the artists explain. "The ability to rate your dealer is radically changing the way people are buying drugs and other controlled items. Buying contraband online means having access to a reliable rating system, while at the same time staying anonymous, all from the comforts of your home."
With projects like this, it's often difficult to ascertain the authenticity of the items and of their provenance. But the artists assure Fast Company that they've taken real precautions. "Together with an arts rights lawyer, we evaluated the work beforehand," they say. "We made a promise to ourselves to show what we get. No censoring. But we do deal with ethical issues every week. Also, because the work involves the people working at the gallery. So we do discuss this issue a lot, but we are taking it one item at a time."
Photo: via Kunst Halle Sankt Gallen, Gunnar Meier
The group show explores other recesses of the deep web. For example, Eva and Franco Mattes's "Emily's Video" is an edited video compilation of volunteer reactions to horrifying footage culled from the dark web. Robert Sakrowski has curated a YouTube history of Anonymous. Seth Price presents a guide to "disappearing" and "dropping out of mainstream society" in America, based on materials he collected from the Internet.
"By exploring the Darknet from an artistic viewpoint we hope to critically evaluate mass surveillance, and to study alternative structures and forms of communicating outside mass surveillance," say the !Mediengruppe Bitnik artists. "How is identity formed in these networks? How is communication and exchange possible in anonymous networks? What forms of trust building arise? How do you trust each other if you don't know to whom you are talking to? How can we as artists examine these questions in a meaningful way?" Their questions are meaningful to everyone, of course—even if not all of us will try to get answers via the purchase of some yellow happy pills.
Ten teams in six countries spanning 13 international time zones worked on Assassin's Creed: Unity—a massive, four-year undertaking that presented the game's maker, Ubisoft Entertainment S.A. of Montreuil, France, with major practical and logistical challenges—and some serious creative and technical hurdles, as well. The video game maker's modern, innovative approach to international collaboration overcame one of its greatest challenges in a fictional French assassin stalking his prey in Paris circa the French Revolution in 1789.
The latest installment of the popular video game series, out this week, is Ubisoft's first foray into the new generation of gaming consoles (alongside Assassin's Creed: Rogue for the previous console generation, with versions of both games available for PCs). The previous six Assassin's Creed top-tier, AAA titles, as they are known in the industry, have sold upward of 78 million units since 2007, according to Ubisoft.
"Assassin's Creed is a big brand, and we definitely needed multiple studios, multiple people working on it," Lesley Phord-Toy, producer at Ubisoft's Toronto studio, said in an interview after her Nov. 10 talk at a Fast Company Innovation Uncensored event at the video game maker's San Francisco facility.
The Montreal studio took the lead on the game, mapping out the vision and providing coordination for nine other units stationed in Canada, France, Ukraine, Romania, China, and Singapore. The overarching goal was to translate that vision into a high-quality game built within budget and on time. That challenge was compounded by a natural conflict: Production, project management, technical direction, and other managerial tasks are all often at odds with the creative process.
"If your role is to create, it can be restrictive, because you also have to deliver. We want to establish a vision without being too focused on delivering," Phord-Toy said.
To help resolve that conflict, Ubisoft divided creative direction from production management to free creators to focus on developing the vision, rather than having to unduly worry about schedules, logistics, and other operational aspects of the project, which were separated into a management function.
Lesley Phord-Toy
"Division of creative and production management became much more important in the previous generation" of game consoles, explained Phord-Toy, who has worked in video games for 12 years and is a founding member of Ubisoft Toronto. "With PlayStation 2, we had teams of 30 to 50 people, but with PlayStation 3, we had teams of hundreds of people. You need processes and structures to support them properly. Dividing the roles minimizes conflict of interest."
Rather than merely decentralizing work across roles and studios, Ubisoft aimed to empower employees to contribute innovative ideas at all levels of the company, as the latest generation of video games continues to require large teams. One of the ways in which Ubisoft implemented this was to ensure that creative directors do not have staff directly reporting to them—i.e., people whom they must manage. Instead, that management function is handled separately by a manager, allowing the creative director to put full attention on the content. Managers generally have seven to 10 people reporting to them, with managers and support at every level, which enables staff to have access to resources they need in their roles, and to collaborate effectively across the organization.
"It gives each team more ownership and less dependency on other studios," Phord-Toy said. At the same time, because they were no longer bogged down by dealing with minutiae, it freed creatives and other employees up to consult and collaborate with each other within studios and with their counterparts at other studios.
In addition, Ubisoft Montreal had a team dedicated to co-productions, coordinating all of the studios working on Assassin's Creed: Unity, and facilitating inter-studio collaboration. "I have an assistant producer in Montreal who I can call," Phord-Toy pointed out. "I encourage my assistant director to communicate with the assistant director at other studios, and my technology lead to collaborate with the technology lead at other studios. So, people in the same roles can help each other," she said.
Ubisoft's approach to encouraging collaboration between studios was put to the test over a third of the way into the flagship game's four-year development process. The Toronto studio, where Phord-Toy works, had over 100 people working on Assassin's Creed: Unity—all of whom they needed when they received a new assignment. Ubisoft Toronto, responsible for building half of 1789 Paris at full-scale from the Left Bank onward, was tasked with enabling four-player cooperative gameplay and developing that content for the game, a challenge that arose from fan demand.
The problem was that the existing narrative vision defined by the lead Montreal studio would be broken by Toronto's multi-player, or "Brotherhood" mode—in which every player was engaged as the game's protagonist Arno Dorian.
The challenge translated into two goals: spur players to work co-operatively, and resolve the "narrative dissonance" caused by four versions of the main character in the same place, at the same time.
Ubisoft Toronto, in consultation with the Montreal studio, addressed these challenges by introducing each Brotherhood mission with a cinematic segment, akin to a miniature movie trailer, narrated by a new female character named The Enabler. The cinematic segments transition the game from action to a scripted segment, introduces the content, frames the history, presents a compelling reason for players to pursue a task together, sets objectives, and transitions back into gameplay. (Historian Laurent Turcot of the University of Québec, Trois-Rivières, guided the development of historical content.)
However, gamers tend to hate cinematic sequences because they can't participate in the events unfolding on their screens. To minimize that pain, Ubisoft Toronto set a one-minute time limit for Brotherhood cinematics.
It was difficult, but Phord-Toy believes that the solution worked, with an added benefit: "It's a sneaky way of us getting a history lesson to our fans."
The solution, which would not have been possible without interdisciplinary collaboration between the Toronto and Montreal teams (as well as some of the other studios), is a prime example of Ubisoft's process at work, according to Phord-Toy. "There are multiple examples—maybe hundreds—where ideas have been shared between studios," she added.
With its operational model proven at such a large scale, Phord-Toy thinks it may may be beneficial in other industries, particularly at the intersection of complex systems and creative content.
Citing movies as a counter-example, she said that from the storyboard phase one can usually envision the final product, with the benefit of smart planning. "Games are more complicated. What's unique to games is the level of iterative development we need to do," Phord-Toy said. "We iterate on technology and features, but also create content on top of that technology.
"You never really know what the gameplay experience will be until you have it in your hands, because you can't really predict what a player will do in the game."
Even as it celebrates its internal workflow successes, Ubisoft's global collaboration model is a living thing that is constantly growing, Phord-Toy noted. "Every studio is incorporating this model, and improving on it with every game Ubisoft is making today," she said.
Selfies are nothing new. But selfies taken by a drone that uses your wrist as a launchpad? That just might be—at least Intel's CEO Brian Krzanich thinks so.
After a 10-month-long competition aimed at helping Intel figure out how wearable computing is going to evolve, the winner is a 40-gram wrist-worn quadcopter that unfolds, flies up, shoots a photo, and boomerangs back to your arm. It's called Nixie, and the team behind it was awarded $500,000.
Nixie, the wearable drone camera.
The founders describe it as "a flying wristband that can give aerial photography to you." Think of it as the futuristic offspring of a GoPro and a drone, or Harry Potter's golden snitch with a built-in gimbal and camera.
"It was the intellectual property about throwing it and, depending on the velocity you throw it at, having it go a certain distance, snap a picture, turn, and come back," Krzanich tells Fast Company. "We thought that was a creative use of technology. It was addressing a market that was growing fast and hungry for it."
Hungry is right. Selfie-sticks are a growing trend, especially outside of the U.S. There is consumer demand for ways to take wider, hands-free, and group shots. And according to 2014 stats by the Camera And Imaging Products Association, digital still camera (DSC) shipment numbers are dropping like a stone—giving way to new options. The market has shifted. According to Krzanich, that's exactly how Nixie convinced judges they were winners.
"The videoclip that they used [in presenting to judges] was one of the YouTube videos that had taken off virally," he explained. "It was a woman in Mexico that had thrown her GoPro 10 times, randomly. She put a delay timer on it, and she threw it up 10 times to snap this wonderful picture of her on the beach. It had a couple million views. The Nixie team said 'Hey, this person was doing our research for us. There's a target audience and a demand for this product.' It was pretty convincing."
The working prototype certainly helped the demo too, and it was built in a sprint. "There were six days between when [Nixie's inventor and founder] Christoph found out about the submission deadline and when we had to submit something," explains project manager Jelena Jovanovic. "I said, 'But we need a working prototype and we need a video and we need a business plan.' And he said, 'Yeah, let's do it.'"
"It was an epic six days," Jovanovic laughs.
The first problem: how to create flexible wristband that bends into a circle but that is still stable enough to fly. Christoph Kohstall, who is not only the inventor but is also an experimental physicist, talked materials and design. "We are using standard 3-D printing (ABS) for the production of the prototype," he said. "ABS doesn't bend, but it's configured kind of like a bike chain." The camera stability, according to Krzanich, comes from a gimbal. "They have a small gimbal in the bottom of the drone that stabilizes it," he shares. Even with all that, the mechanics of getting a steady shot are not easy. Nor are the algorithms Nixie needs to make it boomerang back to you.
"Once we made it to the semi-finals and the finals, we had to incorporate navigation and the camera," says Jovanovic. Some of that technology is already patented. The rest is their secret sauce.
In addition to refining the design, they are sorting out which sensors to use—necessary for things like altitude change and positional movement. "This is a new kind of navigation challenge that wasn't there before," says Kohstall. "There are drones on the market that navigate on GPS. Our drone's gotta be able to do more."
"Our goal is to move really, really quickly because, of course, there are competitors in the space," says Jovanovic adds. "We're just going to stay ahead of the game."
Someone finally found the key to the storage closet where departed Racket would-have-been-founder Matt Taibbi locked Alex Pareene, Edith Zimmerman, and the rest of his staff way back in August. Once released it took them approximately five minutes to create Racket Teen, which is literally the best Tumblr I have ever seen, and I don't mean that merely to damn them with faint praise. With the latest on intelligence-news hotties like Jeremy BAEhill, exclusive Marco Rubio cover pics, and even stylish aspirational sponcon from hot #brand Ralph Lauren Horse Champagne, Racket Teen is what Racket should have been all along. Pareene & co could have launched this, even allowing for very generous weed-smoking breaks, about two days after they were hired, so I can only imagine what kind of nightmarish Event Horizon-type conditions must prevail inside the First Look offices to have prevented them from doing so.
Great Things to Read:Jenn Schiffer's new var t;: Matisse. Jenna Wortham's lovely, humanist "Everybody Sexts." Elmo Keep on the Mars mission that will probably never happen.
I was terribly remiss on Tuesday in forgetting to tab the intern's Vanity Fair interview with Anil Dash. It's quite good. Today he brings us Adrian Chen on Anonymous:
Here's a fact: The internet is kinder to white men. If you've ever spent any time on Reddit, 4chan, or, really, nearly any other site where Issues are Being Discussed, you'll find that the prevailing assumption is that you're a white man; if you deviate from the script by revealing your gender, sexuality, or race, there will always be at least one person who cries that you've only done so for Special Treatment.
Today, Superjournalist and Former Target Of Internet Hate Adrian Chen dives back into that white male id (aged 18-30, natch, an advertiser's holy grail) and surfaces with this beautiful thing in The Nation: "The Truth About Anonymous's Activism". It's a very nice story about the Anonymous "movement" that details exactly how cruel and stupid these people are while wearing the skin of a book review (Gabriella Coleman's a_Hacker, Hoaxer, Whistleblower, Spy: The Many Faces of Anonymous_). Chen also takes on the idea of techno-utopianism—nerds know best, don't you watch the Big Bang Theory???? BAZINGA—calling Anonymous nothing more than the philosophy's latest shit-smelling extrusion.
Similarly, the members of Anonymous barge into issues they know nothing about and proceed with the only logic they understand—believing, as Coleman does, that their position as a technological elite gives them an innate political ability. Along the way, they are helped by a tech-crazed media desperate to find a tech angle in struggles for social justice, like Ferguson.
I can only say a prayer for Adrian's mentions. Enjoy your Special Treatment. RIP.
Bijan's mentions are a thirst-trap 24/7. I have a Tweetdeck column just to watch them.
.@paulcarr Hi Paul, Racket Teen is currently hiring a Social Engagement Editor and we'd love to see your CV — Racket Teen (@RacketTeen) November 13, 2014
Right now, you are drowning in radio waves, satellite beacons, cellphone transmissions, and digital signals. Thankfully, you can't really tell. But what if there was a way to perceive the invisible fields of flowing information that surround us every day?
As the user walks through the city, a hacked iPhone sniffs out nearby Wi-Fi. The interface identifies traceable characteristics like locations and data rates, and translates everything into sound using a pre-programmed language of parameters. The sound is then transmitted to a modified hearing aid, creating sonic representations all around the user.
"On a busy street, we may see over a hundred independent wireless access points within signal range," Jones tells New Scientist. "The strength of the signal, direction, name and security level on these are translated into an audio stream made up of a foreground and background layer: distant signals click and pop like hits on a Geiger counter, while the strongest bleat their network ID in a looped melody," Swain adds.
Screenshot: via Phantom Terrains
You can hear a sample of these sounds on the project website. It's appropriately ambient and electronic. In action, the sounds originate from the direction of the router's location, clicking more frequently as the signal strength increases. There is a particular low sound that signifies the security mode of the network. With enough practice, users will begin to recognize the specific pitches of particular broadcast channels. Like Swain, they would be listening to the "hum and crackle of invisible fields all day" and picking up the "familiar gurgle of the public Wi-Fi hub," all blending into the normally audible.
Think of it as augmented hearing. In his New Scientist article, Swain, who has been losing his hearing since his 20s, explains how sound aids already sort of do this. Unlike glasses, which focus what is blurry, the aids' interface is more complicated. As it picks up the audio emissions around a person, it sorts them into noise and sound, magnifying certain signals and pushing others way into the background.
The Phantom Terrains tool takes what is normally thought of as a prosthetic to the next level. As the creators explain, "By using an audio interface to communicate data feeds rather than a visual one, Phantom Terrains explores hearing as a platform for augmented reality that can immerse us in continuous, dynamic streams of data." Welcome to the age of technologically utilized synesthesia, enhancing human perception to fit our digital age.
If you were daydreaming about building an app that taps into Netflix's huge trove of movies and TV shows, we're sorry to say that your dreams are officially crushed. As of today, the video streaming giant is axing its public API, effectively cutting off developers from accessing its data without a special partnership.
It's not as though devs are being caught off guard. Netflix announced this move in June, so everyone's had plenty of time to come up with a plan B or move onto something different. It's a bummer for the open, mashup-able web. But Netflix is leaving the door open for official partnerships.
The casualties won't be all that widespread: Public API calls make up a tiny percentage of the activity Netflix sees. Most of the requests—about five billion per day, according to GigaOm—come from Netflix's various apps across phones, tablets, connected TVs and other devices with the video service built in.
And in an effort to keep some of the more popular third-party services chugging along, Netflix has entered into formal partnerships with some developers who will be able to access their firehose of content data. So if you're into Fanhattan, Yidio, NextGuide, or Can I Stream It?, there's no need to freak out.
Several other apps and sites will start returning 404 errors as of today. One such unlucky service is Letterboxd, a social network focused on movies. In a blog post, cofounder Matt Buchanan announced that despite their efforts to keep the feature alive, Letterboxd is shutting down its Netflix integration today. To placate frustrated users, Letterboxd is offering refunds to its paid members.
Atlantic Media's Cameroon-based anti-sharing site This period dot cm went into public beta today. This period dot cm, named after a Twitter convention that has already been lampooned to death is a site where members are prevented from sharing all but one of the interesting links they see every day. The site's other innovative counter-viral measures include a name that is impossible to Google and a design that pays clever homage everyone's first Ruby on Rails project, advances which took at least six months to develop. Early users are already hailing it as "a potential Ello-killer!"
Paper Magazine is having an extraordinary week. First there was the butt; now, Julie Klausner got them @Dril. This is a coup, and not only because the tab is hilarious.
For those who don't know, Dril is essentially the ur-Weird Twitter account: whoever's behind it is mysterious and hilarious in roughly equal measure. Which is to say: If you're not already, follow it now. We'll wait.
[...] the dream of the dad is someone who has stopped giving a fuck about all that. Someone who is the model of how not to try, how to just be, who doesn't understand why you'd make fun of him for wearing socks with sandals, who makes you the weird one for thinking it's an issue.
Today's lesson? Get weird.
Dril is literally the only good thing about weird Twitter. Also, for the record, I do not wear socks with sandals.
And they were singing: byeeee bye Netflix public API
I was chasin' after JSON but the river's run dry
And them good old boys better have put catch() after try(),
or this'll be the day that they die()…
Potatoes are legit sexy and I will fight anyone who says otherwise.
Today in Email:Colin Lecher at The Verge FOIAed the FCC's internal emails about John Oliver's net neutrality segment, and they're funny but also check out how fast this FOIA response showed up! Dang. Buzzfeed is banned from WHO info because, you know, not emailing someone is a valid punishment right? PR emailers: just don't.
~But February made me shiver, Despite the hot takes I deliver…~
Today in Tabs has not one but TWO guest tabbers coming up next week! I'm pretty excited. #TabsSeasonTwo is just getting started. Tell your friends to subscribe or read us on FastCoLabs.
How do Detroit techno pioneer Jeff Mills, industrial designer Konstantin Grcic, and sound/visual artist Yoshi Sodeoka see musical instruments 200 years from now? For Future Instruments 2214, a project commissioned by the Red Bull Music Academy, artist Kim Laughton produced immaculate digital renderings based on their ideas, and they're quite fantastical.
Tyler, The Creator jests (we think) that radio will be clogged with screams, the sound of choice for musicians of the era. "Someone will create a small cube box where the artist can scream into it, and it will have every single effect possible to make the scream distinct from other ones," he says in the description.
In Jeff Mills' future, "The Clone 101" is "physical and mental reality player" with a liquid coating comprised of microscopic sensors. It acts as a haptic full body suit and lets the wearer hear and feel the vibrations of the music, as well as experience and sense the psychological state of the music creator.
Seth Woods, cellist and PhD candidate of the Centre for Research in New Music program, imagines a digital exoskeleton of "animated titanium" that creates a sonic landscape by tuning in with the movements of performers, turning gestures into sounds.
The most joyful of the 10 imagined instruments, which includes edge-to-edge screen synthesizers and naturescapes that play themselves, is Yoshi Sodeoka's Zen Sonic Satellite 3000. Orbiting the planet, the satellite constantly gathers data on the "global environmental conditions and mental states of people all over the world" and transmits music "barely audible to the human ear" using "radio microwave technology." In his future it is in tuned with the world and is so spiritually unifying that it will end war, stop crime, reverse environmental damage, and fix everything, ever.
"This idea was inspired by—a popular social palliative since the late decades of the 1900's," Sodeoka explained to Fast Company. "Smooth jazz and classical music have often been played in stressful environments such as elevators, shopping malls, and dentist office waiting rooms. It has been proven that relaxing background melodies have soothing effects on people's minds and spirits. It is possible that the Zen Sonic Satellite 3000 will offer a significant evolution from the arguably limited of today..."
In contrast, Kim Laughton's own idea is beautifully terrifying. In his future, there are no humans—a massive semi-organic landscape covers the Earth, integrated with the artificial intelligence that survives us. All the music that's left is the "flow" of "computer light."
When asked why Sodeoka's zen-full satellite instrument is so optimistic, he said, "I don't know. It's just that all those movies about the future, their vision is all fucked up and stuff. I kind of wanted to try a different angle."
Nancy Duarte has seen a lot of bad tech presentations.
"What happens to most everyone in the technical field is that most of the presentations they draw up are technical documents," says Duarte. "They put a truncated document into PowerPoint."
They rarely tell a good story—you can't do that with a dry set of bullet points read from a screen. A story is full of conflict and contrast. A story has a hero with whom the audience can identify and obstacles which must be overcome.
Duarte insists that a great presenter is not the hero; the audience is. Your job as a speaker is to be a mentor, to show your audience what could be, as opposed to what is, and how they can help create that better world.
Duarte almost followed a technical path herself. "I was a declared math major but I dropped out," she says. "I was very drawn to statistics and math so my work actually expresses itself very mathematically. I got insights from studying Richard Feynman. He was known as the Great Explainer. I think he gives hope to the techie."
In her book Resonate, Duarte analyzes famous speeches using huge infographics of the spoken word. They show how speakers including Martin Luther King Jr. used metaphor, visual words, repetition, and rapid transitions between "what is" and "what could be" to reinforce their point. She is convinced that great speeches contain common patterns, and with a bit of practice, anyone can learn to use them effectively.
Becoming a great presenter doesn't mean that you have to become someone else. "You have to get comfortable being yourself," says Duarte. "Some people try real hard to be someone else. They will try to be very slick and performance-based. But people can tell who your true self is and I would take someone who is just a little bit nervous and geeky over someone who is posing as someone else. So I think you need to amplify the quaintness of your own temperament over trying to put a mask on that feels uncomfortable."
That doesn't mean that you don't have to rehearse. Duarte advises working with a coach. "A coach can teach you things like eye contact, how to look into camera and seem sincere. To me presenting is a bit like the game of golf where you have to have muscle memory, so it's super smart to tape yourself and realize 'Oh, when I make that gesture I look terrible'. I had all these little nervous tics I didn't know I had. What I did is that I closed my eyes and did some of the gestures because I can't observe myself and you have to feel them yourself."
It's all too easy for technical presenters to hide behind technology. They tend to read their slides. And in effect you are putting technology between you and the audience, says Duarte. "If you are going to put the words up then just sit quietly and let them read them and then have an honest conversation about it. They are there to hear you speak, not have your slides speak for you."
Sometimes it's better to have no slides at all. "Let's say you had to do a layoff," says Duarte. "You don't want a bunch of charts to convince people that you did the right thing. You want the page to be blank and you just want to do it in an impassioned, heartfelt apology. You want to fall on your sword and you want to apologize. But if you don't have slides your words need to be pretty visual and colorful."
To stop projecting documents and start telling stories, you need to change the method you use to create presentations. "The way that a research mindset works is very different from a creative mindset," says Duarte. "Go somewhere different. Work in a different environment. Use a different tool like sticky notes. I often use slide sorter view in PowerPoint. It's like sticky notes on a wall. I arrange them and rearrange them and show them to people." She advises finding someone who is a really good communicator and either establishing a formal mentorship or just asking for permission to run your talks by them.
Which presentation tool to use really depends on the type of content you are creating. "Keynote and PowerPoint are chronological whereas Prezi is spatial. Most of the time when we are talking we are doing a chronological talk. With Prezi the reason you would go there is if the meaning is amplified by navigating through space. If you had the entire timeline of humankind and you wanted to zoom into a hundred-year period, then Prezi would be great for that."
Many tech startup presentations, in particular, describe a problem and the speaker's solution. Duarte considers this framework to be too narrow. "Sometimes you have to propose an idea that's not the solution," she says. "It's empowering people to solve that for the future. If you show up and everything's solved—'Here's a problem. I've solved it'—that's off-putting too. With a lot of people, it's only partially solved. They only have a minimal viable product. Software is never done. They only have a little bit done but it's the hope for the future of what they are doing which will bring an investor in. You have to get people to believe that the future you are trying to paint is the right future."
Duarte thinks that the key to a great presentation is plenty of contrast: transitions between what is and what could be, positive and negative emotion, different forms of media and sharing the stage with other people. "It's a form, not a formula. Craft your greatest talk and then go back to see if it has enough contrast in it," she says. "If you are using sticky notes, use a plus sign or a minus sign for what could be and what is. We have a color-coding system also for positively emotionally charged and negatively emotionally charged."
Every time something changes in your presentation the audience will be re-engaged. "So you can have a slide animate or the audio changes so I'm listening to you and then I'm listening to someone else," explains Duarte. "It's a trick. Luring and re-luring them (the audience) by changing it up a lot."
Emotion is one of the most powerful tools in the storyteller's arsenal but it must be used in an appropriate way for your particular audience. "One of the reasons that Benioff (Marc Benioff, the CEO of Salesforce) can get away with a lot of excitement at Dreamforce is that pretty much everyone there does sales and they are a highly emotionally charged, extroverted audience, whereas if you had that same thing at a biotech conference it wouldn't fly."
But deploying the right amount of emotional appeal can be the key to cracking a difficult crowd. "There's a great story in my book about this guy from Stanford, he is a chemist, and he wanted to raise money for his lab," says Duarte. "He had 15 minutes to do his talk and he added a thin veneer of personal story through the whole thing. He won the award above others whose stuff was actually on a par with his. If he had done the whole thing as a story, they wouldn't have gotten it, but even with an analytical type of audience they still have a heart and can be moved. You would just do a thinner layer of that kind of anecdotal storytelling and emotional appeal."
Many techies concentrate on the content at the expense of the visuals, but then so does Duarte, at least initially. "A designer who is trained to visualize my thinking—my art director works on all my books— amplifies the meaning, sometimes changes the meaning of what I am trying to say, because she knows how to make it more clear," says Duarte. "My favorite thing is when I get to co-create with a creative visual thinker. That's the most magical way to do it."
Not everyone has the time or budget to use a designer but Duarte thinks that it is possible to teach yourself to some extent. "Get some of the really beautiful design annuals, design books so that when you are looking at it you think 'I find this beautiful. Why is this beautiful to me?' and you start to see patterns in how the type is done, in the color. There's actually a bit of a science behind design and if you can start to identify the patterns of why decisions were made by a designer, you can actually get a good ways there."
Technical founders and CEOs often put a lot of effort into presentations for an external audience and not enough on their own team. "You are creating a movement," says Duarte. "Look at Dr. King. He spoke as much to his own team that was helping him as he did externally."
"I do my own internal vision talk and I spend about 60 hours on it," Duarte continues. "If I reserve those couple of weeks in January it removes a massive amount of friction all year long because you nail it. It saves you time in the long run. We had a supercritical new move we were going to make where we were going to structure ourselves differently, where everyone was keeping their jobs but everyone was getting a new boss and you would think that would be very unsettling but I got a standing ovation because it was really exciting."
It's no secret that New Yorkers don't think too much about pay phones any more. A quick stroll around the city will reveal that many pay phones don't work and many are just empty booths, lacking actual phones. But the pay phones are a vital piece of city infrastructure, especially in disaster situations. With the need to preserve that infrastructure and the opportunity to reimagine the public terminal, NYC's Department of Information Technology and Telecommunications (DoITT) was tasked with finding a way to preserve pay phones while making them more useful to people in the 21st century.
To accomplish that, DoITT solicited proposals from companies around the world. After a lengthy process, the city has selected a proposal by a consortium of companies collectively called CityBridge. Over the next four to five years, CityBridge will build out what it is calling the LinkNYC network. Each individual terminal will be called a Link and will offer blazing-fast Wi-Fi, touch-screen interfaces, the ability to quickly make 911 and 311 calls, and free charging stations for mobile devices.
The four companies that comprise CityBridge are Titan, Control Group, Qualcomm, and Comark.
Titan is one of the two major operators of current old-school pay phones. The advertising relationships that Titan has built over the years will provide important continuity in the revenue stream for LinkNYC. New Yorkers may know Control Group as the company that has built a handful of interactive information kiosks in the subway system. Naturally, Control Group will be handling the user interface and technology that will power the Links.
Qualcomm will serve primarily in an advisory capacity, building on its experience in wireless connectivity. Comark will do the actual physical manufacturing of the Links.
Acting as a Wi-Fi hotspot is the key feature of the new terminals. LinkNYC will be the fastest free municipal Wi-Fi network in the country. CityBridge boasts that its Gigabit Wi-Fi network would allow a two-hour HD movie to be downloaded in 30 seconds. Users will be able to walk down a street with their phones connecting from one Link to the next without their connection dropping.
Free public Wi-Fi networks are often lackluster, offering slow and frequently interrupted connections. CityBridge plans for LinkNYC to change what public Wi-Fi means for the world.
"One of the reasons why this Wi-Fi network is expected to work where others haven't is because we have linked it to a significant source of revenue, which is the advertising component," says Scott Goldsmith, Titan's chief operating officer.
Between the possible merger of Comcast and Time Warner and uncertain prospects for net neutrality, the future of the Internet remains up in the air. By providing a free, public option for fast Internet, New York City is certainly helping to increase Internet access for all. Representatives of CityBridge specifically state that one of their goals is to help bridge the digital divide "between people of various physical, financial, and technical abilities to connect all New Yorkers to the opportunities that Internet access affords."
"The first payphone was installed in Chicago in 1898 and hasn't changed much since," says New York City Council Member Ben Kallos. "This will revolutionize the structure's design and bring us one step closer to universal broadband in public areas."
The terminals will be built at no cost to New Yorkers, funded entirely through the advertising revenue they will generate. In addition, the advertising is expected to generate $500 million in revenue for the city itself. Thankfully, Links will only have advertising displays in commercial areas, and not on the devices installed on residential streets.
Through the touch-screen interfaces, residents will be able to access city services and get directions their location. Tourists can look up local attractions. As the devices continue to be rolled out and upgraded, CityBridge has plans for Links to be gateways to local businesses. Community members may also be able to engage with issues specific to their neighborhood or block through the Links.
Another local plus: The Link terminals will reportedly be manufactured within NYC's five boroughs, creating over a hundred full-time manufacturing jobs.
Sidewalk real estate is, like all real estate in New York, at a premium. At just 11 inches wide, the terminals will have an incredibly slim profile. In residential areas, the devices will be much smaller without the advertising panels.
Despite their futuristic look and feel, CityBridge has also been careful to pay close attention to accessibility. Representatives for LinkNYC stress that the Links will be fully accessible to people with disabilities.
The sleek design does, however, mean that there's no space for slotting quarters into the machine. But not to worry—the terminals will also offer free calling to all 50 states. No coins required.
