By ROB WALKER
When the Kinect was introduced in November 2010 as a $150 motion-control add-on to Microsoft’s Xbox consoles, it drew attention from more than just video-gamers. A slim, black, oblong 11½-inch wedge perched on a base, it allowed a gamer to use his or her body to throw virtual footballs or kick virtual opponents without a controller, but it was also seen as an important step forward in controlling technology with natural gestures.
In fact, as the company likes to note, the Kinect set “a Guinness World Record for the fastest-selling consumer device ever.” And at least some of the early adopters of the Kinect were not content just to play games with it. “Kinect hackers” were drawn to the fact that the object affordably synthesizes an arsenal of sophisticated components — notably, a fancy video camera, a “depth sensor” to capture visual data in three dimensions and a multiarray microphone capable of a similar trick with audio.
Combined with a powerful microchip and software, these capabilities could be put to uses unrelated to the Xbox. Like: enabling a small drone to “see” its surroundings and avoid obstacles; rigging up a 3-D scanner to create small reproductions of most any object (or person); directing the music of a computerized orchestra with conductorlike gestures; remotely controlling a robot to brush a cat’s fur. It has been used to make animation, to add striking visual effects to videos, to create an “interactive theme park” in South Korea and to control a P.C. by the movement of your hands (or, in a variation developed by some Japanese researchers, your tongue).
At the International Consumer Electronics Show earlier this year, Steve Ballmer, Microsoft’s chief executive, used his keynote presentation to announce that the company would release a version specifically meant for use outside the Xbox context and to indicate that the company would lay down formal rules permitting commercial uses for the device. A result has been a fresh wave of Kinect-centric experiments aimed squarely at the marketplace: helping Bloomingdale’s shoppers find the right size of clothing; enabling a “smart” shopping cart to scan Whole Foods customers’ purchases in real time; making you better at parallel parking.
An object that spawns its own commercial ecosystem is a thing to take seriously. Think of what Apple’s app store did for the iPhone, or for that matter how software continuously expanded the possibilities of the personal computer. Patent-watching sites report that in recent months, Sony, Apple and Google have all registered plans for gesture-control technologies like the Kinect. But there is disagreement about exactly how the Kinect evolved into an object with such potential. Did Microsoft intentionally create a versatile platform analogous to the app store? Or did outsider tech-artists and hobbyists take what the company thought of as a gaming device and redefine its potential?
This clash of theories illustrates a larger debate about the nature of innovation in the 21st century, and the even larger question of who, exactly, decides what any given object is really for. Does progress flow from a corporate entity’s offering a whiz-bang breakthrough embraced by the masses? Or does techno-thing success now depend on the company’s acquiescing to the crowd’s input? Which vision of an object’s meaning wins? The Kinect does not neatly conform to either theory. But in this instance, maybe it’s not about whose vision wins; maybe it’s about the contest.
Theodore Watson bought a Kinect as soon as the gadget was available. He soon acquired 15 more. He admits to a “slight addiction” to the game Call of Duty, but he does not use any of his Kinects to play games. Watson is an artist and a designer who lives in Brooklyn, and his work uses closed-circuit security cameras, graphics cards and gaming hardware “tweaked,” he notes, “for our purposes.”
To use a Kinect with a computer instead of an Xbox, Watson needed a “driver” (basically a bit of software) that did not exist. He joined a small, far-flung, highly dedicated and technically sophisticated community effort dubbed OpenKinect, which sprang up immediately after the Kinect was introduced, to write the code that would make this possible. At the same time, Adafruit, a hobbyist-focused electronics company based in New York, offered $1,000 to the first person or group to write the necessary code in an open-source format.
At the time — this was shortly before the 2010 holiday season — Microsoft’s primary Kinect focus was the mainstream game-playing market. Its first response to OpenKinect seemed predictable: CNET reported an unnamed spokesperson declaring that the company “does not condone the modification of its products” and would “work closely with law enforcement . . . to keep Kinect tamper-resistant.” Adafruit increased its prize, ultimately to $3,000. Within days a developer in Spain posted videos demonstrating that he made his Kinect work with a P.C. OpenKinect refined and spread the open-source driver code, and a variety of “Kinect hacks,” as they came to be called, proliferated in YouTube videos. (An early example involved a Kinect used to create a version of the hand-swipe control contraption Tom Cruise used in “Minority Report.”) Soon Watson and his wife, Emily Gobeille, posted their own video, in which her hand movements were captured by a Kinect and translated onto a screen displaying a computer-generated bird figure, which she controlled like a high-tech puppet.
Watson told me this by phone from Amsterdam, where he and Gobeille had just presented a polished version of their creation as an installation at CineKid, an international entertainment festival, for an audience that included Dutch royalty. The specter of a Microsoft-backed “law enforcement” response to projects like his had obviously faded. In fact, shortly after the open-source driver was finished, one of Microsoft’s top Kinect people appeared on NPR’s “Science Friday” and, in remarks that were widely reproduced across the Web, asserted that OpenKinect participants would “absolutely not” be prosecuted.
In December 2010, Microsoft’s partner PrimeSense, an Israeli company that created the Kinect’s 3-D depth-sensing chip, released its own set of software drivers and code for the so-called hackers to monkey with. A few months later, Microsoft announced it would release its own code kit. It certainly seemed, as Mashable.com put it, that the company had “done a complete 180 when it comes to hacks.”
The idea of a loosely knit band of outsider creative coders forcing a massive company to rethink a crucial new product is appealing. Especially when that company is Microsoft. Fairly or not, Microsoft is widely viewed not as an innovator but as a peddler of me-too products (the Zune, Explorer, Bing, even the look and feel, if you will, of Windows itself) with ruthless business acumen. Open-source zealots point all the way back to 1976, when Bill Gates wrote an “open letter to hobbyists,” complaining that distributors of free, unauthorized copies of software created by what was then called Micro-Soft were disreputable pirates, thwarting progress by removing the financial incentives to improve and develop technology. As John Markoff, a reporter for The Times, put it in his 2005 book, “What the Dormouse Said,” it was a primal ideological standoff between an “anarchic cadre of programmers and hardware tinkerers” and profit-hungry business. That standoff persists.
Last year, I visited the company’s campus in Redmond, Wash., to talk about the Kinect. I was led to an ersatz living room with three couches and a couple of huge TV screens within a mini-mall-style building called the Commons, to meet Alex Kipman. A 32-year-old Brazilian who has spent his entire career at Microsoft, Kipman is credited by the company as the quarterback of the Kinect project; he’s the guy who assured NPR listeners that Microsoft was not siccing lawyers on the OpenKinect community. He has shaggy hair and wore jeans and a mauve T-shirt with a skull design. Three publicists hovered as I floated the notion that Microsoft has a reputation for being a little, you know, control-obsessed.
He cheerily acknowledged that perception and then served up a very different version of the Kinect story. In the hackers’ version, Microsoft had effectively lost control of its own product, thanks to the open-source efforts fueled partly by the Adafruit OpenKinect contest. It had since emerged that Johnny Chung Lee, an employee of Microsoft in the Applied Sciences Group, had covertly bankrolled that competition, and later said — after decamping for Google — that he did so after his “internal efforts” to persuade the company to immediately support the Kinect’s potential beyond gaming went nowhere. “Best $3,000 I ever spent,” he wrote on his blog.
Kipman dismissed the notion that outsiders changed Microsoft’s mind about the Kinect’s potential, or its strategy. For starters, Kinect had not been “hacked” at all, because no one had cracked its proprietary code. “If you’re just person X out there, it’s much more glamorous to call it ‘hacking,’ ” he continued. “From my perspective, it’s ‘tinkering.’ ” Moreover, Kipman waved off Lee’s account of internal resistance to non-Xbox uses for the Kinect: the master plan for the Kinect always included the P.C. “Johnny wasn’t part of any of those conversations,” he told me. “He didn’t even work at Xbox.” The Kinect project attracted lots of enthusiasm within Microsoft, Kipman said, and Lee was a “bright kid” but merely one of “hundreds” of Microsoft coders “contributing from the fringes” who “lacked perspective” on how business works. Besides, he concluded, Microsoft had no problem with the “tinkerer crew” because it anticipated them.
Lee initially declined to comment for this article, but he did e-mail me a response to Kipman’s account: “The tremendous amount of positive press coverage on Kinect projects in the months immediately following launch was worth tens of millions dollars in marketing for Xbox — and remains one of the most culturally interesting aspects of this product, keeping us actively talking about the technology today. Stimulating that for $3,000 seems like good business sense to me.”
Either way, the incident seemed to burnish Microsoft’s reputation. Wired even published an article crediting Microsoft for its forward-thinking attitude toward collaborating with the masses: “No company has made it so easy to hack into a product as popular as the Kinect,” the article asserted, implying that the company planned on home-brew innovators as crucial strategic partners. So I thought I was throwing Kipman a softball when I asked if all the hubbub had ultimately helped Kinect’s highly successful launch.
Not really, Kipman replied. Off-label creators numbered “maybe a thousand,” and millions of Kinect buyers have no awareness of their existence. “We still want to foster” that community, he allowed. “But from the perspective of a multibillion-dollar program, it was neutral. It was neither good nor bad.”
Kinect hackers may not have cared about video games, but what they wanted — a device containing specific high-tech components for just $150 — was achievable specifically because of its connection to something with the scale of the Xbox system. Only a company the size of Microsoft could afford the massive research-and-development costs, and only mass-market appeal could make such a product financially viable.
Kyle McDonald, a digital artist based in Brooklyn, had been working with 3-D sensor technology for years when the Kinect came out, so at first he underestimated the significance of Microsoft’s latest product. But within a week, the hacker videos and online commentary changed his mind, and he bought one. This is evidence of something even more surprising than the possibility that Microsoft had learned to love the hackers: Outsider tech creators have learned to love a Microsoft product. Or if not love, then at least take it seriously. McDonald teaches a class at New York University on “appropriating new technologies.” The goal of any given Kinect hack, he says, isn’t simply to create a high-tech puppet show but to understand how the device works and what its function could be.
The theory that companies should wholeheartedly embrace strange experimentations of people like McDonald turns on a straightforward idea: It’s good for the bottom line. “You get unexpected uses of your products that might contribute to a different direction your company can go,” says Bas van Abel, a designer in Utrecht, the Netherlands, and co-author of the book “Open Design Now.” Established companies may still resist that argument, but more and more upstarts take it for granted that a community of customers, hobbyists and amateurs (or, as van Abel prefers, hackers and artists) will innovate well beyond what any firm can come up with on its own.
According to Henry Chesbrough, a business professor at the University of California at Berkeley and the author of “Open Innovation,” even mainstream companies are starting to agree that tech hobbyists aren’t just consumers but creative partners. This, he says, is purely pragmatic: “We can get more done with less resources by collaborating, cooperating with this community.” He points to Lego’s capitalizing on the unexpected (and decidedly fringy) inventions by users of its Mindstorms motor kits, essentially expanding the company’s place in the mainstream and lucrative education market.
Some at Microsoft do in fact seem pleased about Kinect experimentation from the fringes, particularly those I met in the company’s research division. Microsoft Research, which is separate from the Xbox unit, released the official software developer kit for the Kinect, promoting it with an event called Code Camp last June: a couple of dozen developers, designers, students and others were invited to dream up and create Kinect-using projects in 24 hours, demonstrating the results in a live Webcast. But for the most part, Microsoft’s message about the Kinect in its first year on the market kept it in the predictable context of the living-room entertainment center. In December 2011, the company announced what it called “an all-new Xbox 360 experience,” in which the Kinect’s voice-recognition capabilities acted as a sort of souped-up remote control: you bark “Cate Blanchett movies” at your TV, for example, to see what’s available.
The Microsoft Research Kinect software developer kit — the one announced shortly after the OpenKinect kerfuffle and released last summer — was intended for academics and enthusiasts and carried a license that ruled out business uses; one Kinect hacker complained to me that using it amounted to giving Microsoft free publicity. His point was illustrated just before the Kinect’s first anniversary by an Xbox promotional video, titled “Kinect Effect.” It opens with a product shot, as a narrator explains, over uplifting strings, that “something amazing is happening — the world is starting to imagine things we hadn’t even thought of.” The slick montage features what appear to be actors using the Kinect in various academic, medical and artistic settings. Microsoft hadn’t fought the hackers, but it hadn’t really embraced them either. Instead the company monitored “the Kinect effect” — and appropriated it.
A new version of the Kinect, specifically designed to work with a Windows P.C., came out in February, along with a software kit that would allow developers to create commercial Kinect applications. The P.C. version of the device costs $250, or $100 more than the Xbox version, but the developer software is free. (A slightly upgraded version was released in May.) By March, Microsoft announced team-ups with 350 commercial partners on applications for hospitals, assembly lines, work-force training and so on, including many big corporate names, like American Express and Toyota.
The newer wave of Kinect “hacks” attracting attention in the tech-and-trend press includes interactive shop fronts and billboards, tools for retailers to learn from consumers’ in store-behavior, home security, online banking and something called “natural user interface advertising,” which would use the Kinect to detect, for advertisers, information about who is watching television. Microsoft has also teamed with TechStars, a business incubator, on an effort called Kinect Accelerator, backing a batch of entrepreneurs with business plans built around the Kinect.
In effect, as Tim Carmody, a technology writer, pointed out, the decision to tether market-oriented innovations to a new, P.C.-centric version of the Kinect (and to Microsoft-distributed developer software and commercial-use licensing) has created a separate community of application creation. Creators using the open-source software that makes the original Kinect work with non-Xbox devices can’t really participate in the commercial market — but they’re still free to experiment. That’s not exactly the “open innovation” idea. Then again, it’s also a long way from an attempt to regain full, top-down control over who shapes the way this technology will be used. And maybe this outcome makes sense given the Kinect’s history.
Kyle McDonald, the Brooklyn artist who initially underestimated the Kinect’s importance, still works with the device today. And he’s intrigued by the fact that so many people like him are still experimenting with it, using it in new ways and sharing the code that makes it possible for others to do the same. “People are trying to decide for themselves what they want these technologies to mean,” McDonald says. The answer used to come from defense firms, academia, megacorporations. “It can’t be just them,” he continues. “It has to be everyone.” Maybe that’s supposed to mean a new form of harmony between corporations and outsiders, but in this case, it feels like creative friction.
Posts
No comments:
Post a Comment