"I was sitting on the plane waiting for the flight to take off, and I asked Siri, ‘How long will flight 927 be delayed?’ And Siri came back to me and said the flight would be delayed 15 minutes," recalls Winarsky, who was the SRI executive on the spin-off company’s board before it was sold to Apple. "The guy next to me looked at me and said ‘Wow, I’ve never seen anything like that … why are you in coach?’"
Spoofs and flaws aside, Siri has helped catapult sales of the iPhone 4S and will likely figure prominently into the appeal of the new iPhone 5. Credit is largely due to former Motorola executive Dag Kittlaus, who came on board as the Siri CEO and spotted the market’s hunger for something like her: Smartphones everywhere? Check. Wide availability of fast mobile Internet? Pretty much. The combination of field-based input, tiny screens, and imprecise keyboards? Yep. A tool that took advantage of SRI’s deep expertise in artificial intelligence (gained from decades of work on government projects) paired with strong voice recognition software would prove to be a home run. It was a shining moment for SRI’s innovation model, which insists on identifying the market opportunity for a new invention before making it commercially available.
Winarsky remembers being blown away by how well Siri was integrated into the phone when the 4S came out last year. "[It’s] a spectacular edge that Apple has put forward," he thought. And the cherry on top was Apple sticking with "Siri," a name easily identifiable with the almost 70-year-old company that launched the project.
The smart people who work at SRI International, an organization that’s been involved in cutting-edge research since the end of World War II, aren’t easily surprised. But they never realized the market potential that Apple saw in Siri. As it turned out, brilliant research-oriented inventors used to working for the Department of Defense don’t always make the best innovators.
And even after the Siri score, SRI didn’t exactly kick off a lightning round of consumer product launches, even though SRI is bubbling over with ideas. Today, few products put forth by the organization’s 2,000 employees make it through the gauntlet of SRI’s commercialization board. And only three or four of those are sold, spun off, or licensed by SRI each year. All told, SRI has birthed about a dozen consumer products since Siri. But have you heard of any of them?
Winarsky is quick to point out that the majority of SRI’s budget comes from competing for, winning, and working on government projects. That’s because from its beginnings as Stanford University’s research arm, SRI was happy and honored to work on solving the world’s big problems—curing cancer, developing unmanned aircraft, evading radar detection, renewable energy, robotics, interfacing with computers—and it was deeply invested in an altruistic, nonprofit business model.
SRI worked a lot on "the kinds of things you don’t solve, you just make progress on," according to Bill Mark, a leader on the artificial intelligence projects that laid the foundation for Siri. But even before then, it was the 1980 Bayh-Dole Act that allowed SRI to parlay the inventions and intellectual property that came from its work on government-funded projects into private business opportunities. SRI transformed from a think tank and research organization sustained by government assignments to a center of innovation with its own nest egg. And over the last few decades, it’s helped birth a few important inventions and companies: Nuance Communications, one of the early players in voice recognition; Intuitive Surgical, which championed the idea of doctors using robots to make minimally invasive surgeries more precise; and various cancer treatment drugs. But there has not been a steady stream of Siri-like successes coming out of the campus on Ravenswood Ave. in Menlo Park.
In 2008 SRI unveiled robots that used the principles of static electricity to climb walls. Why didn’t they partner with a toy maker, slap a hard plastic shell on them to make them look like a ladybug, and make the robots the "it" Christmas present of 2010, asks Manu Chatterjee, a Silicon Valley-based strategist with experience at HP and Motorola. "That’s a storyteller problem, right there."
"They’d be hard-pressed to refute the criticism that aside from [those products], they’re not hitting home runs, they’re hitting singles and doubles," Chatterjee says.
Kittlaus had been the storyteller at Siri. As he said in a Chicago Tribune story earlier this year, "(Research) guys are great at solving technical problems. They're not necessarily great at figuring out what the hell to do with it."
While there is a trophy case in SRI’s subdued main lobby populated by mementos from successful acronym projects—DARPA, DOD, and ARPANET and even a couple Emmy awards for work on the technology used in HDTV—it seems secondary, like a happy coincidence, the result of smart people doing great work but not trying to get rich. Even today, with iPhone commercials featuring Siri hobnobbing with a cavalcade of stars—and an SRI program that lets employees share in profits from their commercial successes—SRI lacks the buzz of tech-incubated startups that churn out more millionaires than profitable apps.
During a visit to SRI's campus for a daylong demo, the sheer brainpower on display is impressive. Scientists and researchers—only a few wear white lab coats, the rest are dressed business casual—live in a world where watercooler discussions are less likely to include the names of MVPs in last night's San Francisco Giants game and more likely to reference the official names for cancer-causing proteins (HER-2, the one that causes breast cancer, is a hot topic). The long hallways in some buildings were compared by one researcher to stalls at a flea market. Another ponders a comparison between the lab and the Island of Misfit Toys but says some projects here are hot, live, and destined for consumers. Plenty are considerably cooler.
Janey Ly, a researcher in SRI’s biosciences division who began the laser-aided cancer screening project called FAST at Xerox’s PARC and moved the program down the road to take advantage of SRI’s deep knowledge in her field, seems unfazed by her long timeline. Even working quickly, she and her colleagues won’t see FAST being used widely for at least five years. It took them almost a year to set up their lab on SRI’s campus, and clinical trials (the FDA requires two, one of which has been completed) can take up to three years.
Grit Denker and Rukman Senanayake, key collaborators on the eye-tracking, gesture-following, display-morphing device called bRIGHT, are talking about 10 years before their project might really be finished.
And Rich Mahoney, SRI’s director of robotics and key proponent of the thousand-dollar robotic hand, marvels at how much progress his industry has made making robots that can safely participate in human environments since he was a PhD candidate in the 1980s. (If you shook hands with a robot working on an automaker’s assembly line then, it might rip your arm off.) "Starting in about 1980, every decade was supposed to be the decade of robotics," he says. He’s still patiently awaiting its arrival. "I’ll never say this is the decade of robotics."
There’s a certain amount of audacity required in an organization to believe you have the staff to even tackle these problems, to start from scratch and go through all the R&D growing pains, but there’s no cockiness on display at SRI. Its campus resembles a small liberal arts college while the plans for Apple’s new building evoke the UFO from Close Encounters of the Third Kind. Aside from an updated grill/wok station, the cafeteria probably hasn’t changed in 40 years. And Winarsky, now the vice president of SRI Ventures, an organization with a couple thousand great minds at its disposal, is still sitting in coach.
What Siri has done here, however, is spawn a host of opportunities for SRI to leverage its AI resources and win clients for decades to come. Its aggressively pushing forward with apps like TrapIt, a content aggregator for the iPad; Desti, a search engine focused on travel in the San Francisco Bay area; and the web-banking concierge Lola, developed for Spanish bank BBVA-Compass. Winarsky suggests there will be more to come, with businesses looking to gain a competitive edge in their industry asking SRI to develop an app for them.
"One of the challenges we face is trying to bridge the gap between the kind of things we can do, and the problems people have. So having Siri as a point of reference—‘how good would Siri be at solving your problem, what else would you need?’—provides something people are familiar with," says Mike Freed, an SRI artificial intelligence expert.
But the challenge for SRI and its employees is to maintain the spirit of innovation that swept the campus after Siri’s star turn and push for their projects—some might be world-changing, others might just be fun. Freed, for example, asked during a lunch of fish tacos how some of his nascent projects could make a journalist’s life easier; Mahoney seems committed to being a researcher/entrepreneur and even has a role with a robotics incubator called Redwood Robotics. But what the company might need most are more people like Kittlaus, who could bring storytelling to SRI’s inventions to help turn them into innovations.
Maybe they could make Siri jealous.
[Image: Flickr user Martin Naroznik]