Named “Spin-Off City” by BusinessWeek, SRI International has created and launched more than 60 ventures, with a total market capitalization exceeding $20 billion. Their mission? To create world-changing solutions to make people safer, healthier, and more productive.One innovation to come out of SRI International is Siri, Apple's smart virtual assistant, now a household name. Although one of the best known SRI creations, Siri is just one of many; other acquisitions and spin-offs include Nuance, Tempo AI (acquired by Salesforce), Kasisto, Redwood Robotics (acquired by Google) and more.Nick Triantos is Managing Director of SRI Ventures, where he leads efforts to commercialize cutting edge research from the Information & Computer Science division, through licensing and incubating new companies spanning cybersecurity, bioinformatics, artificial intelligence, machine learning, vision, augmented reality, and speech recognition. Before joining SRI, Nick has been an early team member or founder at 6 startups, an angel investor, a partner at a venture capital firm and Chief Software Architect at NVIDIA. We caught up with Nick, ahead of his presentation at the Virtual Assistant Summit in San Francisco on 28-29 January, to hear more.What is the current mission at SRI International? SRI International is a legendary research institute. Our labs perform research in many fields of science and technology, usually funded by government contracts. To deliver on SRI’s mission of creating world-changing solutions, SRI Ventures finds commercial uses for research that could meet marketplace needs, especially where we could make a major impact. For example, we might develop a navigation system to be used in a military vehicle, and SRI Ventures can then partner with automotive manufacturers, or robot makers, or drone companies, to help them make their products better using the technology we’ve developed. Sometimes, we decide to spin off a whole new company to commercialize the technology, like we’ve done with companies such as Siri, Nuance, and dozens of others. What are the key factors that have enabled recent advancements in virtual assistants? I think there are 3 key factors. First, smartphones have brought us more power-efficient, inexpensive computing platforms. Next, there have been significant improvements in speech recognition algorithms, which have dramatically improved the input to virtual assistant systems. Third, it has become relatively much easier to get access to data through web APIs. If a user asks for the salaries of actors, or the locations of nearby restaurants, or many other kinds of information, those answers can be found relatively easily through web APIs to a myriad of sources. What are the main applications for virtual assistants currently? The largest single use is probably still the built-in assistants on mobile phones, which try to respond to a very broad set of questions. I think the more interesting applications we’re looking at are those where the virtual assistant is quite specialized. These new VPAs can’t answer such a broad base of questions, but their answers tend to be much more in-depth and actionable to the user. For example, Kasisto has built a virtual assistant for consumer banking based on SRI technology. We are also working on virtual assistants in gaming, automotive and other immersive experiences where the virtual assistant can be very sophisticated in its answers. Which new verticals and industries will this expand to in the future? More broadly, I think areas where people today depend upon specialists are all potentially new markets. There are lots of ideas in healthcare and wellness, but also in cybersecurity and other markets where diagnosing a problem and recommending a fix are the reason for the interaction. All of these are great candidates. What developments can we expect to see in virtual assistants in the next 5 years? The current virtual assistants in phones generally wait for a single question, and respond with a single answer. There’s no back-and-forth dialog, which is very unlike an interaction with a human. I expect we’ll see big leaps forward here. I also think personalization needs to play a much greater role in the interaction. When I ask my phone what movie I should see, it should give me a different recommendation than if my daughter asks the same question, for example. What advancements excite you most in the field? Virtual assistant platforms will become available to far more developers so they can invent their own experiences through virtual assistants. I’ve been using computers for nearly my whole life—I’m looking forward to the time when virtual assistants become so good I don’t realize I’m using one. Nick Triantos will be speaking at the RE•WORK Virtual Assistant Summit in San Francisco, on 28-29 January 2016. Other speakers include Dennis Mortensen, x.ai; Deborah Harrison, Cortana; Luca Rigazio, Panasonic Silicon Valley Laboratory; Tim Tuttle, MindMeld; and Dror Oren, Kasisto.
The Virtual Assistant Summit is taking place alongside the Deep Learning Summit.
Tickets for these events are now limited, for more information and to book your place please visit the event page here.