But, he adds, Google co-founder Sergei Brin tells him he’s a better CEO because he says less.

Page lost his left vocal cord over a decade ago after a bad cold and a hoarse voice that never really recovered. Doctors couldn’t explain the problem, never identifying a cause, and assured him that losing the other vocal cord was a virtual impossibility. Fast forward to last year, and again after a cold and a hoarse voice, Page learned that his right vocal cord now had limited movement also.

Both injuries resulted in vocal issues — the first in a slightly weaker voice than normal, and the second with even more impact — but Page says he’s been able to recover, at least partially:

Thankfully, after some initial recovery I’m fully able to do all I need to at home and at work, though my voice is softer than before. And giving long monologues is more tedious for me and probably the audience.  But overall over the last year there has been some improvement with people telling me they think I sound better.

But Page, who says he’s actually feeling quite lucky overall, also says that having paralyzed vocal cords may have made him a better CEO. Why? As co-founder Brin told him, Page is forced to choose his words more carefully … which probably gives him the time to think through what he wants to say before saying it, and also reduces the number of occasions in which thoughtless words cause damage.

I’ve heard that personally on Google earnings calls — Page speaks a little slowly, a lot softly, and with carefully-selected, well-chosen words.

As a result of these challenges, Page came into contact with a Harvard doctor, Steven Zeitels, who is leading a research project on vocal cord paralysis. The problem is relatively rare, he says, and there’s not enough information currently to help doctors learn more about it, and to treat it.

The project is starting with a patient survey to “gather information about the prevalence and extent of vocal paralysis and paresis” from people who have similar problems.

Page’s journey surfaces the question: would every CEO be better by saying less?

Image credit: Google Zeitgeist video

July 9-10, 2013
San Francisco, CA

Tickets On Sale Now

Let’s be frank: Mobile ads suck. Not only are they dull and not making anyone as much money as they should, but people aren’t even clicking them intentionally. In fact, almost forty percent of mobile clicks are made by accident, according to a recent report from eMarketer.

While lots of companies are throwing lots of money at the problem, voice software maker Nuance is looking at it from another, more interesting angle: Instead of focusing on clicks as the success metric for mobile ads, Nuance wants to use voice recognition to engage people in a way that banner ads never could.

“We’re taking the voice assistant tech out of apps and putting it in ads,” Nuance advertisement vice president Mike McSherry said of Voice Ads, Nuance’s new voice-enabled ad platform.

While Nuance’s voice tech is already used in cars, businesses, and hospitals, McSherry says that Voice Ads owes a lot to Siri, Apple’s own voice assistant software. “Siri created an industry. People play with it. It’s witty. Why not give these same sorts of tools to ad agencies?” McSherry said.

Can mobile ads be…entertaining?

At the core of Voice Ads is measurable engagement, which is something that’s been pretty rare for advertising. Think about it this way: For most of its history, advertising has been a one-way street. You watch television and advertisers talk to you via commercials; you read VentureBeat and advertisers serve you banner ads (which can be relevant to you but often aren’t). Absent in both those things, of course, is actual engagement: Advertisers are saying quite a lot to you, but you’re not saying much back in return.

“Magic 8 Ball, will Voice Ads help me make money?”

That idea is turned on its head with Voice Ads, which gives advertisers a tool to interact consumers in a way that’s both direct and, well, entertaining.

To illustrate this somewhat crazy idea, McSherry opened up an interactive magic eight-ball ad for Alpha, a fictional deodorant brand. “Should I buy that ring?” he asks the ad, which, after responding affirmatively  advises McSherry to “use Alpha”, which “smells like money in the bank.” (He did this multiple times, with multiple questions, all of which returned unique responses. Nuance’s tech is definitely robust.)

The conversation was largely silly, but it was still enjoyable to watch – and that’s exactly the point. McSherry wasn’t just watching an advertisement — he was interacting with it.

While Alpha is a hypothetical brand, the sort of engagement that example demonstrates would be ideal for real companies like Geico, which has already built its brand presence around the voice of its gecko mascot. Just imagine an advertisement that allows the Geico gecko to give you personal insurance advice based on your vocal input. (If Geico’s ad guys aren’t already brainstorming something like this, I’d advise them to get to it.)

More engagement, more data

Aside from engagement, Voice Ads also means data. With the platform’s underlying tech, advertisers have a much better idea of which people are actively — and not accidentally – interacting with their ad campaigns.  ”We can measure how much people are actually using this stuff. There are no more accidents,” Nuance marketing chief Peter Mahoney pointed out.

All of these factors combine to support one key takeaway: In Nuance’s ideal world, voice ads will do to advertisements what touch screens have done to computing: If your ads don’t have a voice component, lots of people are going to think there’s something wrong with them — just like people now expect all devices to have touch screens.

Coming to a mobile voice ad near you?

That idea, however, is more of a fantasy right now than reality. Before Nuance answers the expectation question, it first has to answer the adoption one: How many brands and ad agencies will actually end up using Voice Ads?

New technology always takes time to percolate, but McSherry seems confident that once advertisers and brands see what can be done with Voice Ads, it’s going to be tough for them justify avoiding it. “Everyone’s trying to create deeper brand experiences on mobile devices. Voice Ads does that,” McSherry said.

So far things are looking good. Not only has Nuance already inked deals with ad firms Digitas, OMD, and Leo Burnett, but it also has digital advertising companies Millenial, JumpTap and Ad Marvel on board as well. That’s a pretty significant vote of confidence for an experiment that’s just starting out. (McSherry will be providing the first wider public demonstration of the technology at the Mobile Summit today in Sausalito, where VentureBeat has invited 180 leading mobile executives to debate ways to push forward the mobile economy).

The other concern is more relevant to the average smartphone user: How much more annoying will voice-enabled ads be than existing ones? As we’ve seen from banner ads, pop-ups, and behavioral tracking, what’s good for advertisers is often not terribly great for users. While McSherry didn’t have a clear answer for how well the average person is going to respond to Voice Ads,  Nuance obviously hopes mobile users respond well.

My view? Anything that can help companies make money on mobile is ultimately a good thing. Let’s just not make it too annoying — okay, guys?

Photo: Ricardo Bilton/VentureBeat

Google acquired research startup DNNresearch today, after recently awarding the team of three $600,000 for their work in neural networks and language and image processing.

DNNresearch was born at the University of Toronto’s department of computer science. It was founded by Professor Geoffrey Hinton and his two graduate students Alex Krizhevsky and Ilya Sutskever. The company does research on “deep neural networks,” or our language processing centers. Srizhevsky and Sutskever recently created a system “that dramatically improved the state of the art in object recognition,” according to the University of Toronto, making the company an interesting purchase for Google’s search product.

Image and voice queries will likely become more common in the mobile age, as TechVibes points out, and thus improving Google’s processing abilities will make its search even stronger.

Google confirmed the buy to VentureBeat in an email, though it did not provide any more on what the company’s talent will be used for.

The terms of the deal have not been disclosed other than the fact that Hinton will be able to split his time between research projects at the university and those at Google. Krizhevsky and Sutskever will move and focus full-time on Google.

The University of Toronto is otherwise excited to deepen its relationship with Google, likely because the company has already shown its willingness to give money when it’s deserved.

“This is a wonderful opportunity for Geoff, and a great opportunity for the department,” said University of Toronto computer science chair Sven Dickinson in a statement. “In recent years, we have been expanding our industrial relations, and this acquisition represents a wonderful opportunity to strengthen our existing ties with Google, one of the world’s most innovative IT companies.”

Neural connections image via Shutterstock

Sept. 9 - 10, 2013 San Francisco, CA

Early Bird Tickets on Sale

Telecommunications giant Ericson has released a massive report on the state of the mobile world. And clearly, unless you live in the almost-fully-penetrated European and North American markets, everything is up and to the right.

China, the rest of Asia, and Africa together accounted for 75 percent of new mobile subscriptions in the third quarter of 2012, with 38 million in China, 31 million in the rest of Asia, and 25 million in Africa. That contrasts rather sharply with only one million new subscriptions in western Europe and two million in North America.

With 6.4 billion global cellular subscriptions and a world population just shy of seven billion, it would seem that almost everyone has a mobile phone or other mobile device connected to the internet. However, Ericsson, says, those 6.4 billion subscriptions are spread over only 4.3 billion people, meaning that many people have more than one plan … for example, a phone and a tablet, both with a cellular connection.

Which leads to a penetration percentage of well over 100 percent in some areas:

Source: Ericsson

Mobile subscriptions penetration globally

40 percent of all phones sold in the third quarter, Ericsson says, were smartphones, but only 15 percent of the global installed base of phones is smartphones … so there is still a long runway for growth. In fact Ericsson is predicting smartphone subscriptions alone will reach 3.3 billion in 2018.

Source: Ericsson

Global total traffic in mobile networks. Yellow is voice; red is data

Global data traffic is skyrocketing as well — doubling between Q3 2011 and Q3 2012 — while voice traffic is only growing slowly. From perhaps 10 petabytes of monthly mobile data traffic in the first part of 2007, the world’s use of mobile data has exploded to about 900 petabytes, per month, today.

In the same period, voice traffic has only grown from perhaps 80 petabytes to about 190.

By 2018, Ericsson expects mobile traffic to grow 14 times to over 13,000 petabytes of global monthly data, as mobile PCs will gobble up over 10 GB each per month, and smartphones about 2 GB.

Speeds will increase to handle that influx of data traffic, rising to a current global median of 1.3 megabits/second as LTE networks continue to roll out. LTE can already offer download speeds in the 100 Mbps range, on supported devices, and an more-advanced version known as LTE-Advanced will offer peak

data speeds higher than 1 gigabit/second. Only about 100 LTE networks are currently in operation, covering 455 million people. By 2017, Ericsson expects LTE to be available to about half the world’s population.

Interestingly, Ericsson measured the network impact of free versus paid apps as part of the study. Free apps, which are generally, ad-supported, use much more data than premium apps — connecting with the network 30 times more frequently and use over 100 times more data … presumably for ad downloads.

Tethering is also a massive data consumer, especially iPhone and Android users who use data services on their smartphones to get network access on their laptops and tablets. Users who tether use 20 times as much data as non-tethering mobile subscribers.

The full report is available at Ericsson’s website.

Image credits: Ericsson

July 9-10, 2013
San Francisco, CA

Tickets On Sale Now

Adoro falar con xente na súa propia lingua.

You probably don’t understand that, but Google does. Google Voice Search just added 13 new languages, including Norwegian, Swedish, Finnish, and Portuguese to bring its total linguistic competency to 42 languages and dialects.

Meanwhile, Siri knows English, French, German, and Japanese, although Apple will be adding another 11 when iOS 6 hits in the fall.

Adding languages is not easy, but it’s a task that is custom-built for big data, and Google has big data in its back pocket. Before a language comes onstream, the company needs a minimum of hundreds of thousands of phrases, which the company can then parse as its servers learn to translate.

For some languages it’s easier than others: Google says that while Romanian follows predictable pronunciation rules, Swedish does not. (Perhaps that’s why there are so many good Romanian hackers?)

The full list of new languages also includes Basque, Bulgarian, Catalan, Hungarian, Icelandic, Serbian, Slovak, and Galician.

(No, I’d never heard of Galician before either. Wikipedia says it’s the language that is spoken in Galicia, which makes sense, and that Galicia is an autonomous region of northwestern Spain. Another day, another lesson.)

The update is rolling out now but won’t be fully operational for another week. Android users with version 2.2 (Froyo) or later get it automatically; others may need to install the Google Voice app. And iPhone users will presumably also be getting the update as part of the bundle of services in the Google for iPhone app, but there is no word on timing — or even a mention of the iOS app in the Google announcement. Read into that whatever you wish.

So that phrase above? It’s Galician for “I love talking to people in their own language.”

Of course, knowing that Google Voice Search and Google Translate aren’t perfect, especially with new languages … with my luck, it probably says something like “Your mother has hairy legs and smells like a goat.”

photo credit: woodleywonderworks via photo pin cc

Clearly, cloud platforms are not reality show contestants, so perhaps I should explain a little bit further.

This type of technology enables developers to integrate multiple methods of communication into their applications via application programming interfaces (APIs). Rather than relying on phones, external networks, hardware, and carriers, companies can bypass the middlemen by incorporating voice and messaging capabilities in to their products through a few simple lines of code.

It is a powerful tool for companies that require in-app communication to conduct business. Telecommunications hardware can be expensive to set up and difficult to maintain, and building it on the web saves money and provides more flexibility to scale the product.

The most well-known player in this space is Twilio. This crowd favorite has raised $33.7 million and has recently expanded its text-messaging platform into 150 countries.

In the arena of voice communication, however, a upstart competitor is zooming ahead.

Since its launch last month, Y Combinator startup Plivo has provided cloud-based local phone numbers in more than 50 countries worldwide, across the US, Europe, South, America, Africa and Asia. In contrast, Twilio’s  voice API — unlike its SMS product — is only available in 12 countries and has yet to make a dent in the Asian or African market.

This rapid progress is largely attributable to the Plivo founders’ combined 17 years of experience working in the international telecom industry. Furthermore, their technology operates independently from geography so it can easily be applied anywhere and everywhere.

The company, founded by Venkatesh B and Michael Ricordeau in 2010, is based in Walnut, Calif. and Bangalore, India, and employs six people.

In many countries, setting up a phone solution is challenging and involves negotiations with carriers, hardware, and network configuration. By focusing primarily on its international presence and leveraging its technology in places like Vietnam or Chile, Plivo hopes to dominate the global arena when it comes to voice communication.

As the rise in texting-related car accidents has shown, there’s a big demand for technology that allows drivers to be connected without distracting them from the road.

Nuance, makers of Dragon Naturally Speaking, may just have that solution. Its new Dragon Drive automotive platform offers drivers the ability to do anything from send text messages and get directions, to search for music and read newspaper articles — all while keeping their eyes on the road. Think of it as Siri on wheels.

The key component is Dragon Drive’s reliance on natural language technology, which allows users to interact with it in a simple, natural way. Say “I’m looking for a chop shop” or ask “What’s the quickest way to the the border” and Dragon Drive will respond naturally with directions.

“Dragon Drive understands what you mean,” Nuance said in a release.

Not new to the car or voice technology in general, Nuance’s products have had applications in fields as varied as construction and health care.

Last June, the company announced that it had acquired SVOX, an in-car voice solutions provider. A clear precursor to today’s news, the acquisition gave Nuance access to SVOX’s stable of  tech, much of which is behind Dragon Drive.

Watch the video below for a better idea of how Dragon Drive works.

For now, this new feature relies on Chrome web browser functionality. From the browser, users can simply click the small, blue microphone icon in the search bar, begin speaking and see results accordingly.

The speak-to-search function has also been available in Google Maps mobile interfaces for a while, since hands-free directions are pretty important to on-the-go mobile users.

“Using voice search can make it easier to find hard-to-spell places (like Poughkeepsie or Liechtenstein) or simply get directions without typing (for example, say ‘Directions from Los Angeles to San Francisco’),” wrote Google engineer Jed Burgess on the company blog.

Here’s an overview of how Google voice search works:

Google’s voice search has been around for a while. The company first launched voice search for iPhone back in 2008.

Those mobile technologies made their way to the desktop earlier this summer at Google’s Inside Search event, where the company announced that people could use their computer’s microphone to accomplish the same search tasks with voice alone.

Last summer, the company launched voice actions for its Froyo-running Android phones. Also, at the end of last year, Google said it was making voice search smarter by bringing personalized voice profiles to its voice search app for Android.

July 9-10, 2013
San Francisco, CA

Tickets On Sale Now

Saygent, a company the uses voice response and analysis to figure out customers emotions, today announced it has secured a first round of funding for $1 million. The funding will be used to hire engineers and secure new customers.

The company targets Fortune 500 companies, like Comcast and the Royal Bank of Canada. It offers the ability to reach a large set of customers via telephone and analyze voice responses through crowdsourcing techniques. The results are meant to help companies figure out audience emotions, like sentiment and intent, towards things like products, services, or customer service. Thus, helping to market easier to these consumers without the expenses of a call center.

As noted by TechCrunch, investors include 500 Startups, Innovation Endeavors, Juvo Capital, Kapor Capital, Kima Ventures, Orefa Investment, PG Ventures, Ty Danco and Matthew Grodin.

The Mountain View, California-based company, founded in 2009, is lead by co-founder Guy Hirsch, a serial entrepreneur who has founded three companies to date, as well as co-founder and chief technologist Mariya Genzel, an MIT graduate.