Roberto Pieraccini: Siri points to a need for voice standards (m-pulse recap)

Another week, another UNTETHER.tv m-pulse episode with Rob Woodbridge and Peggy Anne Salz. Peggy has written another thorough episode recap over on MobileGroove, and once again, the comments of this episode’s special guest are worth calling out specifically.

Last week, Bill Meisel, stated that Siri indicated a paradigm shift and explored the opportunity for mobile voice in the enterprise space. This week, Roberto Pieraccini, author of The Voice in the Machine, speaks to what is necessary to unleash the true power of voice.

Like Meisel, Roberto believes that Siri demonstrates natural language is a “powerful tool” for expressing complicated requests and ideas. However, he believes that present day technology only takes us so far, and current limitations have to be balanced against consumer expectations.

But the intelligence of search services like Siri can’t hide the fact that the industry still has to solve issues around background noise, accents and the sheer complexity of our spoken language. “Siri is extremely beautiful work and engineering,” Roberto observes. “But try to say something complex and Siri won’t understand it.” This is why Roberto’s work as Director of the International Computer Science Institute (ICSI) is about building more sophisticated models that unleash the true power of voice.

Right now, Roberto tells us there is “a new energy in speech research,” with everyone wanting to build a Siri-like app. Siri is fun, and the star of popular culture and comedy skits, but our admiration of voice enabled service could fuel market hype. If future voice services fail to entertain and deliver, then we could see the advance of “another winter of voice recognition,” when we should all be enjoying a great summer.

The solution? Roberto believes that convincing the major voice players to agree on a set of standards is necessary to accelerate innovation and create a profitable market for the burgeoning industry.

Once we have standards the way will be clear for companies to choose where they want to play (and win). Voice is complex and we’ll want apps tailored to deliver us what we need. Watch a movie and want to ‘ask’ the TV who the actor is — and use that as a jumping off point for an entertainment focused search/assistance experience. Yes — there WILL have to be an app for that.

Interested in learning more? Read Peggy’s full write up, or watch the m-pulse episode in question below:

Episode #6

Recorded: February 16, 2012

Can’t see the video? Click here

This episode is brought to you by:

Smartr Contacts
Xobni, Smartr Inbox, and Smartr Contacts automatically find all the people with whom you’ve ever exchanged emails, calls, or SMS messages. They instantly provide a full view of each contact, complete with their photo, job title, company details, and email history — as well as updates from LinkedIn, Facebook and Twitter.




About the author

Douglas Soltys

Douglas is the former Editor-In-Chief of Inside BlackBerry, BlackBerry Cool, and QuicklyBored, which he launched as a mobile gaming industry site. His knowledge of mobile and social media led him to a job at RIM (BlackBerry), where he got to travel the world and do lots of cool things. He is often left-handed, but rarely sinister.

/* ]]> */