The tree versus the shadow

Character is like a tree and reputation like its shadow. The shadow is what we think of it; the tree is the real thing. —Abraham Lincoln

When I was looking for investors for my startup in 1997, I was a lot more interested in the values and character of my potential investors than in the name of the firm on the door. My first investors were not from Sand Hill Road or Palo Alto, but were people I knew would support my company (and me) in the ways we most needed help.

As an entrepreneur, you want to know if your investors will be there for you, help you do the right things, and encourage you to persist and evolve when things seem dark. Perhaps most of all, you want to know they’ve been there before, and that they’ll be calm and confident when things go wrong—when giving up sounds like a pretty good option.

At Google Ventures, we try to be the kind of hands-on investors that quietly help to build companies. We try to be the kind of investors we sought as entrepreneurs ourselves. Starting a company can be a lonely business, and it helps to know you’ve got someone on your side who has been through it before.

Today, we're launching a new Google Ventures blog as an experiment, hoping it will help you get to know the people and companies at Google Ventures a little better. We’ll share how we think about things and why, if you’re an entrepreneur, you might want to talk to Joe, Rich, Kevin and the rest of the team.

Posted by Bill Maris, Managing Partner, Google Ventures

(Cross-posted from the Google Ventures Blog)

Speech Recognition and Deep Learning

Posted by Vincent Vanhoucke, Research Scientist, Speech Team

The New York Times recently published an article about Google’s large scale deep learning project, which learns to discover patterns in large datasets, including... cats on YouTube!

What’s the point of building a gigantic cat detector you might ask? When you combine large amounts of data, large-scale distributed computing and powerful machine learning algorithms, you can apply the technology to address a large variety of practical problems.

With the launch of the latest Android platform release, Jelly Bean, we’ve taken a significant step towards making that technology useful: when you speak to your Android phone, chances are, you are talking to a neural network trained to recognize your speech.

Using neural networks for speech recognition is nothing new: the first proofs of concept were developed in the late 1980s(1), and after what can only be described as a 20-year dry-spell, evidence that the technology could scale to modern computing resources has recently begun to emerge(2). What changed? Access to larger and larger databases of speech, advances in computing power, including GPUs and fast distributed computing clusters such as the Google Compute Engine, unveiled at Google I/O this year, and a better understanding of how to scale the algorithms to make them effective learners.

The research, which reduces the error rate by over 20%, will be presented(3) at a conference this September, but true to our philosophy of integrated research, we’re delighted to bring the bleeding edge to our users first.

--

1 Phoneme recognition using time-delay neural networks, A. Waibel, T. Hanazawa, G. Hinton, K. Shikano and K.J. Lang. IEEE Transactions on Acoustics, Speech and Signal Processing, vol.37, no.3, pp.328-339, Mar 1989.

2 Acoustic Modeling using Deep Belief Networks, A. Mohamed, G. Dahl and G. Hinton. Accepted for publication in IEEE Transactions on Audio, Speech and Language Processing.

3 Application Of Pretrained Deep Neural Networks To Large Vocabulary Speech Recognition, N. Jaitly, P. Nguyen, A. Senior and V. Vanhoucke, Accepted for publication in the Proceedings of Interspeech 2012.

A new way to visualize the global arms trade

Did you know that 60 percent of all violent deaths are due to small arms and light weapons? Small arms, such as revolvers, assault rifles and light machine guns, and ammunition represent a multi-billion dollar industry, and three quarters of the world’s small arms lie in the hand of civilians—more than 650 million civilian arms. As part of the Google Ideas initiative on illicit networks, we’ve created an interactive data visualization of global small arms and ammunition trading to better understand and map the global arms trade.

The tool was produced by Google’s Creative Lab team in collaboration with the Igarape Institute. More than 1 million data points on imports and exports of small arms, light weapons and ammunition between 1992 and 2010 and across 250 states and territories across the world were provided by the Peace Research Institute Oslo (PRIO) small arms database. The visualization reveals patterns and trends in imports and exports of arms and ammunition across the world, making it easy to explore how they relate to conflicts worldwide. More of the implications of this data are discussed in this video from the INFO summit hosted by Google Ideas last month:



You can explore these data points by zooming in and out of the globe, clicking on any country to readjust the view, and using the histogram tool at the bottom to see trading patterns over the years. You can see, for example, that the scale of the global trade in ammunition rivals the scale of trade in actual weapons, an insight underexplored by policymakers today in conflict prevention and resolution.

We built this visualization using the open source WebGL Globe on Google’s Chrome Experiments site; since it is open sourced, we hope to see others use the globe as a platform for bringing other complex datasets to life.

Update 8/16/12: This post has been updated to reflect more accurate numbers.

Posted by Scott Carpenter, Deputy Director, Google Ideas

Giving you a better Google

We work every day to create a more seamless, beautiful user experience—to give you a better, easier-to-use Google. This means continuously improving the products we offer, and recognizing when users of one product might have a better experience with another. Over the past year, we’ve made changes to around 50 products, features and services—donating, merging and shutting things down so we can focus on the high-impact products that millions of people use, multiple times a day. Today, we’re announcing a few more changes:

• We introduced Google Apps for Teams in 2008 to allow people with a verified business or school email address to collaborate using non-email applications from Google like Google Docs, Google Calendar and Google Talk. Over time, we realized that Google Apps for Teams was not as useful for people as we originally anticipated. Beginning September 4, 2012, we'll start converting existing Google Apps for Teams accounts into personal Google Accounts, and shutting down Google Apps for Teams. This change does not affect other editions of Google Apps.
• We launched Google Listen through Google Labs in August 2009, to give people a way to discover and listen to podcasts. However, with Google Play, people now have access to a wider variety of podcast apps, so we’ve discontinued Listen. People who have already installed the app can still use it, but after November 1, podcast search won’t function. You can access your podcast subscriptions in Google Reader in the “Listen Subscriptions” folder and download them from the Import/Export tab.
• Google Video for Business is a video hosting and sharing solution that allowed Google Apps for Business and Google Apps for Education customers to use video for internal communication. Starting this fall, we’ll migrate all videos hosted on Google Video for Business over to Google Drive, which has similar storage and sharing capabilities. All migrated videos will be stored for free and will not count against a user’s Google Drive storage quota.
• Finally, Google maintains 150+ blogs and other communications channels about our products and services, and so over time we'll also be closing a number of Google-created blogs that are either updated infrequently, or are redundant with other blogs. This doesn't mean that we'll be sharing any less information—we'll just be posting our updates on our more popular channels.

Technology has the power to change people’s lives. But to make a difference, we need to carefully consider what to focus on, and make hard decisions about what we won’t pursue. This enables us to devote more time and resources giving you products you love, and making them better for you.

Posted by Max Ibel, Director of Engineering

Our roots go global: Apply for a RISE Award today

The application for the 2013 Google RISE (Roots in Science and Engineering) Awards is now open. Given once a year, Google RISE Awards are designed to promote and support education initiatives in two key areas: Science, Technology, Engineering, Mathematics (STEM) and Computer Science (CS). Google grants awards of $5,000 - $25,000 USD to organizations around the world working with K-12 and university students in these fields.

This year, our community welcomed more than two dozen organizations from around the world, from Denmark to Uganda and California to Romania. The RISE grants have helped these groups to scale their reach by allowing for more scholarship recipients, to deepen their impact by providing hands on robotics kits, and to ultimately inspire their students by creating a community for CS outreach.

Now, the RISE Awards have expanded to include applicants from Latin America and the Asia Pacific region, bringing our total to six continents and 243 countries. All eligible nations are listed on our website.

The growth of technology is undoubtable, and the impact technology will have on our future is equally undeniable. We believe it’s our duty to support students who have the uncanny ability not only to consume technology, but also to create it. We believe that inspiring the next generation of computer scientists will enrich the lives of not only individual students, but also the communities and countries they live in.

Show us what you can do to get students excited about STEM and CS! Submit your application by September 30, 2012. Awardees will be announced by January 2013.

Posted by Roxana Shirkhoda, K-12 Education Outreach Specialist

Reflections on Digital Interactions: Thoughts from the 2012 NA Faculty Summit

Posted by Alfred Spector, Vice President of Research and Special Initiatives

Last week, we held our eighth annual North America Computer Science Faculty Summit at our headquarters in Mountain View. Over 100 leading faculty joined us from 65 universities located in North America, Asia Pacific and Latin America to attend the two-day Summit, which focused on new interactions in our increasingly digital world.

In my introductory remarks, I shared some themes that are shaping our research agenda. The first relates to the amazing scale of systems we now can contemplate. How can we get to computational clouds of, perhaps, a billion cores (or processing elements)? How can such clouds be efficient and manageable, and what will they be capable of? Google is actively working on most aspects of large scale systems, and we continue to look for opportunities to collaborate with our academic colleagues. I note that we announced a cloud-based program to support Education based on Google App Engine technology.

Another theme in my introduction was semantic understanding. With the introduction of our Knowledge Graph and other work, we are making great progress toward data-driven analysis of the meaning of information. Users, who provide a continual stream of subtle feedback, drive continuous improvement in the quality of our systems, whether about a celebrity, the meaning of a word in context, or a historical event. In addition, we have found that the combination of information from multiple sources helps us understand meaning more efficiently. When multiple signals are aggregated, particularly with different types of analysis, we have fewer errors and improved semantic understanding. Applying the “combination hypothesis,” makes systems more intelligent.

Finally, I talked about User Experience. Our field is developing ever more creative user interfaces (which both present information to users, and accept information from them), partially due to the revolution in mobile computing but also due in-part to the availability of large-scale processing in the cloud and deeper semantic understanding. There is no doubt that our interactions with computers will be vastly different 10 years from now, and they will be significantly more fluid, or natural.

This page lists the Googler and Faculty presentations at the summit.

One of the highest intensity sessions we had was the panel on online learning with Daphne Koller from Stanford/Coursera, and Peter Norvig and Bradley Horowitz from Google. While there is a long way to go, I am so pleased that academicians are now thinking seriously about how information technology can be used to make education more effective and efficient. The infrastructure and user-device building blocks are there, and I think the community can now quickly get creative and provide the experiences we want for our students. Certainly, our own recent experience with our online Power Searching Course shows that the baseline approach works, but it also illustrates how much more can be done.

I asked Elliot Solloway (University of Michigan) and Cathleen Norris (University of North Texas), two faculty attendees, to provide their perspective on the panel and they have posted their reflections on their blog.

The digital era is changing the human experience. The summit talks and sessions exemplified the new ways in which we interact with devices, each other, and the world around us, and revealed the vast potential for further innovation in this space. Events such as these keep ideas flowing and it’s immensely fun to be part of very broadly-based, computer science community.

Celebrate the Kennedy Space Center’s 50th birthday with Street View

For fifty years, NASA's Kennedy Space Center in Florida has been the launch point for a generation of space technology and exploration. Countless enthusiasts (including this one) grew up longing to see a space shuttle up close and walk in the paths of astronauts. Today, a collaboration between NASA and Street View is enabling people around the world to take a trip to the doorway to outer space, and see Kennedy as it transitions into a multipurpose launch complex for the next 50 years of space innovation. This location is our largest special collection of Street View imagery to date, totaling 6,000 panoramic views of the facilities, and expanding our mission to document the world’s most amazing places.

Street View fleet at Kennedy Space Center

Among the images you can now explore online with the click of your mouse are the space shuttle launch pad, Vehicle Assembly Building and Launch Firing Room #4. Gaze down from the top of the enormous launch pad, peer up at the towering ceiling of the Vehicle Assembly Building (taller than the Statue of Liberty) and get up close to one of the space shuttle’s main engines, which is powerful enough to generate 400,000 lbs of thrust. From these vantage points, anyone can live out his or her childhood dream of becoming an astronaut.

Atlantis Orbiter in the Vehicle Assembly Building - View Larger Map

Saturn V rocket on display in the Apollo/Saturn V center - View Larger Map

Even though they recently entered retirement, you can still get an up-close, immersive experience with two of the Space Shuttle Orbiters—the Atlantis and Endeavour —with Street View.

We’d like to thank NASA for making this project possible and giving all of us the chance to digitally walk in the shoes of all of the pioneering astronauts, scientists, engineers and technicians that made our space dreams possible. To explore more of our largest special collection yet, visit our Street View gallery at maps.google.com/nasa.

Posted by Ryan Falor, Product Manager, Google Street View

Google+ Hangouts and the NFL: fantasy football, face-to-face-to-face

The 2012 NFL season is just over a month away, which means one thing for millions of fans around the world: it’s time for fantasy. Fantasy football has been around since the 1960s, and has become a great way to follow your favorite athletes—and stay in touch with the people you care about. This year, alongside the NFL, we're giving fantasy football fans a new way to connect throughout the season: Google+ Hangouts. Here’s how it works in a nutshell:

• From any NFL.com/fantasyfootball page—including the Draft and Game Centers—you can now click a single button to start a hangout with your league
• Later this month, you’ll see a live indicator on NFL.com/fantasyfootball if anyone in your league is already hanging out

Because most fans select their fantasy rosters in real time, and many leagues consist of people from across the globe, we think you'll find that hangouts are a natural extension of your (healthy) obsession. Not only can you connect live with your friends on draft day; you can actually gloat celebrate in person, regardless of where your league members are located. We're also upping the number of hangout participants from 10 to 12 on NFL.com/fantasyfootball, since the standard fantasy league has a dozen players.

Fantasy football hangout (top window), live indicator on NFL.com (right), chat bar on NFL.com (bottom-left)

Fantasy football hangouts don’t stop at the draft, however. Throughout the season you’ll be able to:

• Hang out with your friends to conduct trades, review league business or just chat while watching the game on TV
• Enjoy fantasy hangouts on the go or on the couch using the new Google+ app for phones and tablets (Android, iOS)
• Get advice from top NFL Fantasy commentators as they hang out with fans and discuss the week’s results

The +NFL Network will broadcast weekly highlights from some of these commentator hangouts. And of course, you can always share fantasy articles and tips with your circles—either from the NFL’s site, or their newly-launched Google+ page.

Visit NFL.com/fantasyfootball today to sign up, and to prep for your draft via hangout. If you like football but haven’t yet played fantasy, this year’s a great time to start. After all, hangouts help make fantasy football about more than just the game; they’re a great way to build lasting relationships on any given Sunday (and every day in between).

Posted by Nikhyl Singhal, Google+ Product Director

Natural Language in Voice Search

Posted by Jakob Uszkoreit, Software Engineer

On July 26 and 27, we held our eighth annual Computer Science Faculty Summit on our Mountain View Campus. During the event, we brought you a series of blog posts dedicated to sharing the Summit's talks, panels and sessions, and we continue with this glimpse into natural language in voice search. --Ed

At this year’s Faculty Summit, I had the opportunity to showcase the newest version of Google Voice Search. This version hints at how Google Search, in particular on mobile devices and by voice, will become increasingly capable of responding to natural language queries.

I first outlined the trajectory of Google Voice Search, which was initially released in 2007. Voice actions, launched in 2010 for Android devices, made it possible to control your device by speaking to it. For example, if you wanted to set your device alarm for 10:00 AM, you could say “set alarm for 10:00 AM. Label: meeting on voice actions.” To indicate the subject of the alarm, a meeting about voice actions, you would have to use the keyword “label”! Certainly not everyone would think to frame the requested action this way. What if you could speak to your device in a more natural way and have it understand you?

At last month’s Google I/O 2012, we announced a version of voice actions that supports much more natural commands. For instance, your device will now set an alarm if you say “my meeting is at 10:00 AM, remind me”. This makes even previously existing functionality, such as sending a text message or calling someone, more discoverable on the device -- that is, if you express a voice command in whatever way feels natural to you, whether it be “let David know I’ll be late via text” or “make sure I buy milk by 3 pm”, there is now a good chance that your device will respond how you anticipated it to.

I then discussed some of the possibly unexpected decisions we made when designing the system we now use for interpreting natural language queries or requests. For example, as you would expect from Google, our approach to interpreting natural language queries is data-driven and relies heavily on machine learning. In complex machine learning systems, however, it is often difficult to figure out the underlying cause for an error: after supplying them with training and test data, you merely obtain a set of metrics that hopefully give a reasonable indication about the system’s quality but they fail to provide an explanation for why a certain input lead to a given, possibly wrong output.

As a result, even understanding why some mistakes were made requires experts in the field and detailed analysis, rendering it nearly impossible to harness non-experts in analyzing and improving such systems. To avoid this, we aim to make every partial decision of the system as interpretable as possible. In many cases, any random speaker of English could look at its possibly erroneous behavior in response to some input and quickly identify the underlying issue - and in some cases even fix it!

We are especially interested in working with our academic colleagues on some of the many fascinating research and engineering challenges in building large-scale, yet interpretable natural language understanding systems and devising the machine learning algorithms this requires.

Sparking a better conversation with Wildfire

Businesses around the world—from neighborhood restaurants to major retailers—are embracing social media to share information and forge stronger relationships with their fans and customers. We know because we are one of those businesses—on Google+, Twitter, Facebook, Pinterest, YouTube and LinkedIn.

Indeed, a social presence can complement all marketing campaigns—search, display, video, mobile, offline ads and more.

With this in mind, today we’re happy to share that the Wildfire team will be joining Google. Their co-founders, Victoria Ransom and Alain Chuard, launched their startup just four years ago. Since then, they and their team have built a service that helps businesses like Virgin, Cirque du Soleil, Gilt Group and Spotify manage their social efforts across numerous social websites. It’s a platform for brands to manage their pages, apps, tweets, videos, sponsorships, ads, promotions and more, all in one place.

The ultimate goal is better and fresher content, and more meaningful interactions. People today can make their voices heard in ways that were previously impossible, and Wildfire helps businesses uphold their end of the conversation (or spark a new one).

In a complex and changing landscape, businesses want to manage and measure these efforts in an integrated way. We’ve been working towards this end for some time. For example, Google Analytics helps businesses measure the contribution of hundreds of social sites; our Admeld service has helped to serve ads in Facebook developers’ social apps; and our DoubleClick platform enables clients to run and measure ads across social websites. On Google+, brands use services like Vitrue, Buddy Media and others to manage their pages, with many more to come.

With Wildfire, we’re looking forward to creating new opportunities for our clients to engage with people across all social services. We believe that better content and more seamless solutions will help unlock the full potential of the web for people and businesses.

Update August 15, 2012: Our acquisition of Wildfire has now closed.

Posted by Jason Miller, Product Management Director