Posted by Joanna Smith, Developer Advocate and Giles Hogben, Google Privacy Team
Marshmallow introduced several changes that were designed to help your app look after user data. The goal was to make it easier for developers to do the right thing. So as Android 6.0, Marshmallow, gains traction, we challenge you to do just that.
This post highlights the key considerations for user trust when it comes to runtime permissions and hardware identifiers, and points you to new best practices documentation to clarify what to aim for in your own app.
With Marshmallow, permissions have moved from install-time to runtime. This is a mandatory change for SDK 23+, meaning it will affect all developers and all applications targeting Android 6.0. Your app will need to be updated anyway, so your challenge is to do so thoughtfully.
Runtime permissions mean that your app can now request access to sensitive information in the context that it will be used. This gives you a chance to explain the need for the permission, without scaring users with a long list of requests.
Permissions are also now organized into groups, so that users can make an informed decision without needing to understand technical jargon. By allowing your users to make a decision, they may decide not to grant a permission or to revoke a previously-granted permission. So, your app needs to be thoughtful when handling API calls requiring permissions that may have been denied, and about building in graceful failure-handling so that your users can still interact with the rest of your app.
The other aspect of user trust is doing the right thing with user data. With Marshmallow, we are turning off access to some kinds of data in order to direct developers down this path.
Most notably, Local WiFi and Bluetooth MAC addresses are no longer available. The getMacAddress() method of a WifiInfo object and the BluetoothAdapter.getDefaultAdapter().getAddress() method will both return 02:00:00:00:00:00 from now on.
However, Google Play Services now provides Instance IDs, which identify an application instance running on a device. Instance IDs provide a reliable alternative to non-resettable, device-scoped hardware IDs, as they will not persist across a factory reset and are scoped to an app instance. See the Google Developer's What is Instance ID? help article for more information.
User trust depends largely on what users see and how they feel. Mishandling permissions and identifiers increases the risk of unwanted/unintended tracking, and can result in users feeling that your app doesn’t actually care about the user. So to help you get it right, we’ve created new documentation that should enable developers to be certain that their app is doing the right thing for their users.
Growing up, my parents were daily reminders of the sacrifices made by earlier generations of Black Americans to give people like me the opportunities they were denied. To this day, their stories propel me to continue the fight for justice. I am far from alone—reflecting on a shared history inspires millions around the world to work toward equality. But without some record, those stories and the passion they ignite could get lost.
Artworks, artifacts and archives have the power not only to give a story life, but to encourage action and incite change. That’s why the Google Cultural Institute is excited to add records from institutions like the Smithsonian National Museum of African American History and Culture, the Studio Museum and Amistad Research Center and many more—bringing together important archives from Black history for anyone to access not only during Black History Month, but throughout the year.
From the New Orleans Jazz Orchestra to the historical records of Frederick Douglass and Dr. Martin Luther King Jr., this collection includes 26 new institutions (50 overall) contributing 5,000+ items and more than 80 curated exhibits. It includes new Street View imagery and three Google Expeditions, including an exploration of the resurgence of Jazz in New Orleans with Irvin Mayfield and Soledad O’Brien. You can see a 360 degree YouTube video made in conjunction with that Expedition here:
In The Baltimore Museum of Art’s exhibition “Questioning the Canon,” you can see Mickalene Thomas’s Le déjeuner sur l'herbe: Les Trois Femmes Noires and compare it side-by-side with the Manet original to see the ways Thomas has subverted the subject-matter of this canonical white European work.
You can trace along the paths of history by reading Frederick Douglass’ letter to his former master, and read the original manuscripts of Dr. King’s ”I Have a Dream” and “I’ve Been to the Mountaintop” speeches. Absorb Dr. King’s personal letter to wife Coretta Scott King at the beginning of his four-month prison term for non-violent protest, then cut to photographs documenting his momentous first handshake at the White House with President Lyndon B. Johnson. Collecting these works into one place provides unprecedented access to a vital part of history that is too often forgotten. By comparing works of art and texts of speeches to find commonalities and distinctions, we can also build on the past to inspire ourselves and others. And while today is the first day of Black History Month, the work of remembering our history is necessary year round—which is why these records will be there on the Cultural Institute for generations to come.
Posted by Valeisha Butterfield Jones, Head of Black Community Engagementhttp://1.bp.blogspot.com/-GQmKUucIo2M/Vq7wbvb29_I/AAAAAAAAR0I/tC4JD65XndA/s1600/Screenshot%2B2016-01-29%2Bat%2B4.02.49%2BPM.pngValeisha Butterfield JonesHead of Black Community EngagementAUTHOR TEAM
Like all good projects, this one started out because we had an itch to scratch…
As Site Reliability Engineers who manage corporate infrastructure at Google, we deal with a large number of internally used services that need to be load balanced for scalability and reliability. In 2012, two different platforms were used to provide load balancing, both of which presented different sets of management and stability challenges. In order to alleviate these issues, our team set about looking for a replacement load balancing platform.
After evaluating a number of platforms, including existing open source projects, we were unable to find one that met all of our needs and decided to set about developing a robust and scalable load balancing platform. The requirements were not exactly complex - we needed the ability to handle traffic for unicast and anycastVIPs, perform load balancing withNAT andDSR (also known as DR), and perform adequate health checks against the backends. Above all we wanted a platform that allowed for ease of management, including automated deployment of configuration changes.
One of the two existing platforms was built uponLinux LVS, which provided the necessary load balancing at the network level. This was known to work successfully and we opted to retain this for the new platform. Several design decisions were made early on in the project — the first of these was to use theGo programming language, since it provided an incredibly powerful way to implement concurrency (goroutines and channels), along with easy interprocess communication (net/rpc). The second was to implement a modular multi-process architecture. The third was to simply abort and terminate a process if we ended up in an unknown state, which would ideally allow for failover and/or self-recovery.
After a period of concentrated development effort, we completed and successfully deployed Seesaw v2 as a replacement for both existing platforms. Overall it allowed us to increase service availability and reduce management overhead. We're pleased to be able to make this platform available to the rest of the world and hope that other enterprises are able to benefit from this project. You can find the code athttps://github.com/google/seesaw.
Today, we’re rolling out the AdWords app on iOS to all AdWords customers globally – you can download it from the App Store. With the AdWords app, many campaign activities can now be managed while you’re on the go from the convenience of your iPhone:
Monitor campaign performance like clicks, CTR and CPC
Update bids and budgets
Act on suggestions that may help improve campaign performance
Get real-time alerts and notifications about your billing and ad status
Call a Google expert
Customers like The Honest Company, MuleSoft, and PMG use the AdWords app to easily manage their campaigns, stay in touch with the needs of their customers, and quickly access important business insights – from anywhere.
“Amidst the hectic holiday festivities, this app saved me from having to leave the dinner table to monitor performance and make quick changes to my accounts. That meant more time with my family. I'm excited for what's to come!” – Josh Franklin, Manager, Search Marketing, The Honest Company
“The app helps me access high level data on the go which can come in handy in the boardroom, or anytime I need to quickly understand how our campaigns are performing. Also, having the ability to make adjustments to our campaigns – such as changing bids and budget – is invaluable.” – Nima Asrar Haghighi, Director, Digital Marketing & Analytics, MuleSoft
“The consumer shift to mobile means our retail clients' campaigns have to be responsive to meet the needs of consumers at all times of the day. The app makes it easy for us to address issues without being chained to our laptops. PMG has been able to deliver prompt account adjustments from campaign to keyword level for our clients, as well as keep our customer satisfaction rates high.” – Kyle Knox, Account Manager, PMG
When Google was a few years old, we wrote up a list of Ten things we know to be true. The list includes items like “Focus on the user and all else will follow” as well as “Fast is better than slow.” It would be tough to say that much of the mobile web has adhered to these principles. Users often get frustrated by poor experiences in which sites load slowly or will lock up trying to load resources that clog their data connections.
Today, the AMP team announced the launch of an analytics component that will enable measurement on AMP pages. The Google Analytics team is committed to helping our users measure their content wherever it appears. So, for publishers looking to use AMP to provide an improved user experience, we’ve released Google Analytics measurement capabilities for Accelerated Mobile Pages. AMP support in Google Analytics makes it easy to identify your best content and optimize your user experience.
How Google Analytics Support Works
How to Get Started
Before you get started with AMP Analytics, you’ll need to get started with AMP itself. The AMP website contains a great introduction to getting started. Once you have an AMP page up, it’s time to start thinking about how you’d like to measure its performance.
We recommend that you use a separate Google Analytics property to measure your AMP pages. AMP is a new technology that’s going to mature over time. As such, some of the functionality that you’re used to in web analytics won’t immediately be available in AMP analytics right away. AMP pages can appear in multiple contexts, including through different syndication caches. Because of that, a single user that visits an AMP version of a page and a HTML version of a page can end up being treated as two distinct users. Using a separate Google Analytics property to measure AMP pages makes it easier to handle these issues.
Multiple technology partners, including Google Search, Twitter, Pinterest, and LinkedIn have announced that they’ll start surfacing AMP pages in the coming months. The Google Analytics team is excited to support AMP from day one and look forward to growing our offering as AMP’s capabilities expand.
Posted by Dan Cary, Product Manager and Avi Mehta, Software Engineer
Did you know that there are more mobile devices than people? In 2014, there were 7.4BN* devices and an estimated 7.2 billion people. Mobile is quickly becoming the most popular way to get online, even exceeding the time spent on desktop computers.** This has contributed greatly to the growth of the app ecosystem, with consumers spending on average 37 hours a month*** on mobile apps. As apps are becoming more prominent in our everyday lives, it’s more important now than ever to learn app development.
That’s why we’re launching the second AdMob Student App Challenge, an app building competition open to students around the world. If you’re a student who loves to code or has a great idea for a mobile app, this is your chance to build an app, learn how to make money from it, and win awesome prizes. The prizes will include a week-long trip to San Francisco and a visit to the Googleplex.
To win the challenge, you’ll need to build a great app (either Android or iOS) and create a sound business plan that shows how you managed the project, gained users, and leveraged AdMob to make money. The contest will be judged by a panel of app industry leaders. To learn more the judges, visit here.
The winning team will receive:
An all expenses paid, week-long trip to San Francisco, including a visit to the Googleplex in Mountain View, CA
A featured spot on the Google Play Store
A promotional video starring you and your app
A healthy amount of Google schwag, including a new Google device
Last year’s global winner was Phani Gaddipati, who created Stacks Flashcards, an app that lets people create electronic flashcards on any subject, quiz themselves and analyze their performance. Check out his video to learn more.
In addition to one overall winner, the finalist from each of the four global regions (North America, Latin America, Europe Middle East & Africa, and Asia Pacific) will receive Google devices and see their app featured on the AdMob website.
The challenge starts today, and you’ll have until June 28, 2016 to build your winning app. Be sure to visit the AdMob website to learn more and register. Follow us on AdMob G+ and Twitter and keep an update on #AdMobSAC16 too, for regular updates on the challenge. .
Best of luck -- we can’t wait to see what you build!
Posted by Henry Wang
Product Marketing, AdMob
*Cisco, Cisco Visual Networking Index: Global Mobile Data Traffic Forecast Update 2014–2019 White Paper, February 2015 **eMarketer, US Time Spent with Media, October 2014 ***Nielsen, Smartphones: So Many Apps, So Much Time, July 2014
We launched our Vulnerability Reward Program in 2010 because rewarding security researchers for their hard work benefits everyone. These financial rewards help make our services, and the web as a whole, safer and more secure.
With an open approach, we’re able to consider a broad diversity of expertise for individual issues. We can also offer incentives for external researchers to work on challenging, time-consuming, projects that otherwise may not receive proper attention.
Last January, we summarized these efforts in our first ever Security Reward Program ‘Year in Review’. Now, at the beginning of another new year, we wanted to look back at 2015 and again show our appreciation for researchers’ important contributions.
2015 at a Glance
Once again, researchers from around the world—Great Britain, Poland, Germany, Romania, Israel, Brazil, United States, China, Russia, India to name a few countries—participated our program.
Here's an overview of the rewards they received and broader milestones for the program, as a whole.
Android Joins Security Rewards
Android was a newcomer to the Security Reward program initiative in 2015 and it made a significant and immediate impact as soon as it joined the program.
We launched our Android VRP in June, and by the end of 2015, we had paid more than $200,000 to researchers for their work, including our largest single payment of $37,500 to an Android security researcher.
New Vulnerability Research Grants Pay Off
Last year, we began to provide researchers with Vulnerability Research Grants, lump sums of money that researchers receive before starting their investigations. The purpose of these grants is to ensure that researchers are rewarded for their hard work, even if they don’t find a vulnerability.
We’ve already seen positive results from this program; here’s one example. Kamil Histamullin a researcher from Kasan, Russia received a VRP grant early last year. Shortly thereafter, he found an issue in YouTube Creator Studio which would have enabled anyone to delete any video from YouTube by simply changing a parameter from the URL. After the issue was reported, our teams quickly fixed it and the researcher was was rewarded $5,000 in addition to his initial research grant. Kamil detailed his findings on his personal blog in March.
Established Programs Continue to Grow
We continued to see important security research in our established programs in 2015. Here are just a few examples:
Tomasz Bojarski found 70 bugs on Google in 2015, and was our most prolific researcher of the year. He found a bug in our vulnerability submission form.
You may have read about Sanmay Ved, a researcher from who was able to buy google.com for one minute on Google Domains. Our initial financial reward to Sanmay—$ 6,006.13—spelled-out Google, numerically (squint a little and you’ll see it!). We then doubled this amount when Sanmay donated his reward to charity.
We also injected some new energy into these existing research programs and grants. In December, we announced that we'd be dedicating one million dollars specifically for security research related to Google Drive.
We’re looking forward to continuing the Security Reward Program’s growth in 2016. Stay tuned for more exciting reward program changes throughout the year.
Updated versions of the Google Docs, Sheets, and Slides Android apps are now available on Google Play. These versions include the following new features:
Import and export additional file formats in Google Docs, Sheets, and Slides - As long as you’re online, you can now import and export the following file formats in and from the Docs, Sheets, and Slides Android apps:
Import - .txt, .rtf, and .html (in Docs)
Export - .docx, .rtf, .pdf, .txt., and .html (from Docs), .xlsx, .pdf, .html, .csv, and .tsv (from Sheets), and .pptx, .pdf, and .txt (from Slides)
Open CSV and TSV files in Google Sheets - In the latest version of the Android app for Sheets, you can open, view, and edit spreadsheets that are formatted as .csv or .tsv files.
Read from right to left in Google Sheets - If you use a right-to-left language (e.g. Hebrew) in Sheets, you’ll now see an option in the Android app to format your spreadsheet in that same direction. Once enabled, your columns will progress from right to left (starting with “A”), and your row numbers will appear on the right-hand side of your screen.
Filter data in Google Sheets - You can now apply new filters to spreadsheets in the Sheets Android app, as well as view and change existing ones. (NOTE: This feature became available in the app’s previous release.)
Check out the Help Center articles below for more information.
Launching to both Rapid release and Scheduled release
Gradual rollout (potentially longer than 3 days for feature visibility)
Work with others, with ease. Today’s launches make it super simple to comment on your colleagues’—and your own—Google Docs, Sheets, and Slides files on the web and mobile. Check out what’s new below:
Instant comments in Google Docs on the web - Highlight text or hover over the edge of a page to surface a small commenting icon, then click on that icon to quickly add a comment to that area of the document.
Instant mentions in Google Docs, Sheets, and Slides on the web, Android, and iOS - Start typing someone’s name or email address in a comment, and a list of suggested contacts will appear. Select anyone from that list to trigger an email notification to them—there’s no need to type “+” or “@” before their name!
Comments in the Google Sheets and Slides apps for Android and iOS - Add new and reply to existing comments in the Sheets* and Slides mobile apps—just like you can in the Docs apps for Android and iOS.
Comment swiping in the Google Docs, Sheets, and Slides apps - Simply swipe to move from one comment to the next in the Docs, Sheets, and Slides iOS apps, as well as the Slides Android app (you can already do this in the Docs and Sheets Android apps!).
Try out these new features in Google Docs, Sheets, and Slides, and collaborate with fewer clicks!
Launch Details Release track: Mobile features - Launching to both Rapid release and Scheduled release
Instant mentions in Google Docs, Sheets, and Slides on Android and iOS
Comments in the Google Sheets and Slides apps for Android and iOS*
Comment swiping in the Google Docs, Sheets, and Slides apps for iOS
Web features - Launching to Rapid release, with Scheduled release coming on February 10th
Instant comments in Google Docs on the web
Instant mentions in Google Docs, Sheets, and Slides on the web
*NOTE: Comments in the Google Sheets Android app launched on November 12th, 2015.
Rollout pace: Full rollout (1–3 days for feature visibility)
When Jim, one of the engineers on the Google Slides team, brought a zucchini chocolate cake into the office last week, we knew we had to get the recipe. So we asked him and his wife, Alison, to let us in on the family secret—just in time for Chocolate Cake Day. They worked together in Slides (mobile commenting across Google Docs just launched today!) to perfect the recipe. Alison writes:
Growing up, my grandma made zucchini chocolate cake often, especially when there was a surplus of zucchinis at the local farmer’s market. The cake is ridiculously moist and pairs well with many different frostings, though cream cheese is my favorite.
Thanks to mobile commenting, Jim and I went back and forth on the recipe—Jim on his Nexus 9, me on my iPhone—until we had it just right: Check out our family recipe in Slides. We call it Straka’s Zucchini Chocolate Cake—in honor of my grandma.
Happy Chocolate Cake Day, from our family to yours.
Posted by Alison Zoll, chemist, baker and wife of Jim Zoll, Slides engineer
The game of Go originated in China more than 2,500 years ago. Confucius wrote about the game, and it is considered one of the four essential arts required of any true Chinese scholar. Played by more than 40 million people worldwide, the rules of the game are simple: Players take turns to place black or white stones on a board, trying to capture the opponent's stones or surround empty space to make points of territory. The game is played primarily through intuition and feel, and because of its beauty, subtlety and intellectual depth it has captured the human imagination for centuries.
But as simple as the rules are, Go is a game of profound complexity. There are 1,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000,000 possible positions—that’s more than the number of atoms in the universe, and more than a googol times larger than chess.
This complexity is what makes Go hard for computers to play, and therefore an irresistible challenge to artificial intelligence (AI) researchers, who use games as a testing ground to invent smart, flexible algorithms that can tackle problems, sometimes in ways similar to humans. The first game mastered by a computer was noughts and crosses (also known as tic-tac-toe) in 1952. Then fell checkers in 1994. In 1997 Deep Blue famously beat Garry Kasparov at chess. It’s not limited to board games either—IBM's Watson [PDF] bested two champions at Jeopardy in 2011, and in 2014 our own algorithms learned to play dozens of Atari games just from the raw pixel inputs. But to date, Go has thwarted AI researchers; computers still only play Go as well as amateurs.
Traditional AI methods—which construct a search tree over all possible positions—don’t have a chance in Go. So when we set out to crack Go, we took a different approach. We built a system, AlphaGo, that combines an advanced tree search with deep neural networks. These neural networks take a description of the Go board as an input and process it through 12 different network layers containing millions of neuron-like connections. One neural network, the “policy network,” selects the next move to play. The other neural network, the “value network,” predicts the winner of the game.
We trained the neural networks on 30 million moves from games played by human experts, until it could predict the human move 57 percent of the time (the previous record before AlphaGo was 44 percent). But our goal is to beat the best human players, not just mimic them. To do this, AlphaGo learned to discover new strategies for itself, by playing thousands of games between its neural networks, and adjusting the connections using a trial-and-error process known as reinforcement learning. Of course, all of this requires a huge amount of computing power, so we made extensive use of Google Cloud Platform.
After all that training it was time to put AlphaGo to the test. First, we held a tournament between AlphaGo and the other top programs at the forefront of computer Go. AlphaGo won all but one of its 500 games against these programs. So the next step was to invite the reigning three-time European Go champion Fan Hui—an elite professional player who has devoted his life to Go since the age of 12—to our London office for a challenge match. In a closed-doors match last October, AlphaGo won by 5 games to 0. It was the first time a computer program has ever beaten a professional Go player. You can find out more in our paper, which was published in Nature today.
What’s next? In March, AlphaGo will face its ultimate challenge: a five-game challenge match in Seoul against the legendary Lee Sedol—the top Go player in the world over the past decade.
We are thrilled to have mastered Go and thus achieved one of the grand challenges of AI. However, the most significant aspect of all this for us is that AlphaGo isn’t just an “expert” system built with hand-crafted rules; instead it uses general machine learning techniques to figure out for itself how to win at Go. While games are the perfect platform for developing and testing AI algorithms quickly and efficiently, ultimately we want to apply these techniques to important real-world problems. Because the methods we’ve used are general-purpose, our hope is that one day they could be extended to help us address some of society’s toughest and most pressing problems, from climate modelling to complex disease analysis. We’re excited to see what we can use this technology to tackle next!
Posted by Demis Hassabis, Google DeepMindhttp://3.bp.blogspot.com/-UM6zXm-cXW4/VqkFrP32nlI/AAAAAAAARzw/HmxeOsYvvqk/s1600/Go-game_hero.jpgDemis HassabisGoogle DeepMind
Posted by David Silver and Demis Hassabis, Google DeepMind
Games are a great testing ground for developing smarter, more flexible algorithms that have the ability to tackle problems in ways similar to humans. Creating programs that are able to play games better than the best humans has a long history - the first classic game mastered by a computer was noughts and crosses (also known as tic-tac-toe) in 1952 as a PhD candidate’s project. Then fell checkers in 1994. Chess was tackled by Deep Blue in 1997. The success isn’t limited to board games, either - IBM's Watson won first place on Jeopardy in 2011, and in 2014 our own algorithms learned to play dozens of Atari games just from the raw pixel inputs.
But one game has thwarted A.I. research thus far: the ancient game of Go. Invented in China over 2500 years ago, Go is played by more than 40 million people worldwide. The rules are simple: players take turns to place black or white stones on a board, trying to capture the opponent's stones or surround empty space to make points of territory. Confucius wrote about the game, and its aesthetic beauty elevated it to one of the four essential arts required of any true Chinese scholar. The game is played primarily through intuition and feel, and because of its subtlety and intellectual depth it has captured the human imagination for centuries.
But as simple as the rules are, Go is a game of profound complexity. The search space in Go is vast -- more than a googol times larger than chess (a number greater than there are atoms in the universe!). As a result, traditional “brute force” AI methods -- which construct a search tree over all possible sequences of moves -- don’t have a chance in Go. To date, computers have played Go only as well as amateurs. Experts predicted it would be at least another 10 years until a computer could beat one of the world’s elite group of Go professionals.
We saw this as an irresistible challenge! We started building a system, AlphaGo, described in a paper in Nature this week, that would overcome these barriers. The key to AlphaGo is reducing the enormous search space to something more manageable. To do this, it combines a state-of-the-art tree search with two deep neural networks, each of which contains many layers with millions of neuron-like connections. One neural network, the “policy network”, predicts the next move, and is used to narrow the search to consider only the moves most likely to lead to a win. The other neural network, the “value network”, is then used to reduce the depth of the search tree -- estimating the winner in each position in place of searching all the way to the end of the game.
AlphaGo’s search algorithm is much more human-like than previous approaches. For example, when Deep Blue played chess, it searched by brute force over thousands of times more positions than AlphaGo. Instead, AlphaGo looks ahead by playing out the remainder of the game in its imagination, many times over - a technique known as Monte-Carlo tree search. But unlike previous Monte-Carlo programs, AlphaGo uses deep neural networks to guide its search. During each simulated game, the policy network suggests intelligent moves to play, while the value network astutely evaluates the position that is reached. Finally, AlphaGo chooses the move that is most successful in simulation.
We first trained the policy network on 30 million moves from games played by human experts, until it could predict the human move 57% of the time (the previous record before AlphaGo was 44%). But our goal is to beat the best human players, not just mimic them. To do this, AlphaGo learned to discover new strategies for itself, by playing thousands of games between its neural networks, and gradually improving them using a trial-and-error process known as reinforcement learning. This approach led to much better policy networks, so strong in fact that the raw neural network (immediately, without any tree search at all) can defeat state-of-the-art Go programs that build enormous search trees.
These policy networks were in turn used to train the value networks, again by reinforcement learning from games of self-play. These value networks can evaluate any Go position and estimate the eventual winner - a problem so hard it was believed to be impossible.
Of course, all of this requires a huge amount of compute power, so we made extensive use of Google Cloud Platform, which enables researchers working on AI and Machine Learning to access elastic compute, storage and networking capacity on demand. In addition, new open source libraries for numerical computation using data flow graphs, such as TensorFlow, allow researchers to efficiently deploy the computation needed for deep learning algorithms across multiple CPUs or GPUs.
So how strong is AlphaGo? To answer this question, we played a tournament between AlphaGo and the best of the rest - the top Go programs at the forefront of A.I. research. Using a single machine, AlphaGo won all but one of its 500 games against these programs. In fact, AlphaGo even beat those programs after giving them 4 free moves headstart at the beginning of each game. A high-performance version of AlphaGo, distributed across many machines, was even stronger.
This figure from the Nature article shows the Elo rating and approximate rank of AlphaGo (both single machine and distributed versions), the European champion Fan Hui (a professional 2-dan), and the strongest other Go programs, evaluated over thousands of games. Pale pink bars show the performance of other programs when given a four move headstart.
It seemed that AlphaGo was ready for a greater challenge. So we invited the reigning 3-time European Go champion Fan Hui — an elite professional player who has devoted his life to Go since the age of 12 — to our London office for a challenge match. The match was played behind closed doors between October 5-9 last year. AlphaGo won by 5 games to 0 -- the first time a computer program has ever beaten a professional Go player. AlphaGo’s next challenge will be to play the top Go player in the world over the last decade, Lee Sedol. The match will take place this March in Seoul, South Korea. Lee Sedol is excited to take on the challenge saying, "I am privileged to be the one to play, but I am confident that I can win." It should prove to be a fascinating contest!
We are thrilled to have mastered Go and thus achieved one of the grand challenges of AI. However, the most significant aspect of all this for us is that AlphaGo isn’t just an ‘expert’ system built with hand-crafted rules, but instead uses general machine learning techniques to allow it to improve itself, just by watching and playing games. While games are the perfect platform for developing and testing AI algorithms quickly and efficiently, ultimately we want to apply these techniques to important real-world problems. Because the methods we have used are general purpose, our hope is that one day they could be extended to help us address some of society’s toughest and most pressing problems, from climate modelling to complex disease analysis.
Since the launch of Google Photos, we’ve had a lot of questions around what this means for the future of Picasa. After much thought and consideration, we’ve decided to retire Picasa over the coming months in order to focus entirely on a single photo service in Google Photos. We believe we can create a much better experience by focusing on one service that provides more functionality and works across mobile and desktop, rather than divide our efforts across two different products.
We know for many of you, a great deal of care has gone into managing your photos and videos using Picasa—including the hours you’ve invested and the most precious moments you’ve trusted us with. So we will take some time in order to do this right and provide you with options and easy ways to access your content. We’ve outlined below some of the changes you can expect.
Picasa Web Albums If you have photos or videos in a Picasa Web Album today, the easiest way to still access, modify and share most of that content is to log in to Google Photos, and all your photos and videos will already be there. Using Google Photos, you can continue to upload and organize your memories, as well as enjoy other great benefits like better ways to search and share your images.
However, for those of you who don’t want to use Google Photos or who still want to be able to view specific content, such as tags, captions or comments, we will be creating a new place for you to access your Picasa Web Albums data. That way, you will still be able to view, download, or delete your Picasa Web Albums, you just won’t be able to create, organize or edit albums (you would now do this in Google Photos).
One thing to make clear is that none of this is happening today—if you have a Picasa Web Album you can keep using it as normal. We’ll start rolling out these changes on May 1, 2016.
Desktop application As of March 15, 2016, we will no longer be supporting the Picasa desktop application. For those who have already downloaded this—or choose to do so before this date—it will continue to work as it does today, but we will not be developing it further, and there will be no future updates. If you choose to switch to Google Photos, you can continue to upload photos and videos using the desktop uploader at photos.google.com/apps.
Finally for developers, we will also be retiring some functions of the Picasa API. Developers can learn more here.
Again, none of these changes are happening today, and we’ll continue to update you along the way. We apologize for any inconvenience this transition causes, but we want to assure you that we are doing this with the aim of providing the best photos experience possible. Google Photos is a new and smarter product, that offers a better platform for us to build amazing experiences and features for you in the future.
In our personal lives, we write the perfect message in a card and attach it to the perfect gift. In our professional lives as search marketers, we try to write that perfect message in our ad copy and attach it to the right landing page, but this hasn’t always been easy. That’s why we are excited to share our latest enhancements to DoubleClick Search which will help you quickly find the ad copy and landing page that your customers will love – this Valentine’s Day and beyond.
Find the perfect message
Using search, 50% of consumers lookup brands they’ve never purchased2. Being there with useful information can help you win the love of new customers. Ad copy testing in DoubleClick Search helps you test new ad copy in order to make sure your campaigns are having their greatest impact. Ad copy testing works by normalizing the variables that can impact campaign performance such as flight dates, device type, and display frequency, and measuring the performance of what matters most – the ad copy.
Testing ad copy in DoubleClick Search requires just three quick steps:
Step 1: Write your new ad copy
Step 2: Choose a test goal such as click-through rate, action rate or revenue.
Step 3: Measure results.
When your test is over, you can make changes to your campaigns directly from the testing interface.
“Ad copy testing is a task all SEM managers know to be absolutely necessary and a pillar of a successful account. Previous to DoubleClick Search ad copy testing, we would have to rely on in-house scripts or expensive and clunky 3rd party tools to get the job done. Now that we can take direct action inside of DoubleClick Search, we’re able to save several hours every week putting together and reporting on our testing data and use that time more effectively to help strategically drive our clients’ business forward.” -Kyle Petzinger | Lead, Paid Search, iProspect
With so much last minute shopping, brands that find ways to be useful to consumers in the moments when they are seeking information for their purchases will win. So, when a consumer clicks one of your search ads, which page should you show? A promotional page? A product details page?
A DoubleClick Search landing page test helps answer this question by randomly splitting ad traffic among multiple landing pages on your site. For AdWords engine accounts, we’ve added the ability to specify the device(s) consumers used to reach your landing page in order to ensure you have the most actionable insights about your page performance, no matter what device a consumer used to get there.
We’re excited to see DoubleClick Search customers use the mobile landing page test to optimize the journey for their customers from mobile search to mobile landing page to action.
To learn how to start a landing page test, visit the Help Center.
Meeting all your search marketing needs
Ad copy testing and improved landing page testing in DoubleClick Search are just two of the many ways we are enhancing DoubleClick Search to help save you time and boost your performance. If these features make a difference for your business, let your account manager know. We’d love to hear about your success.
Posted by Amit Varia Product Manager, DoubleClick Search
1 Google Trends, January 2013-December 2015, United States. 2 Google, “The Role of Mobile on the CPG Purchase Journey” study, United States, September 2015. n=371
The latest versions of the Google Docs, Sheets, and Slides Android apps include several new features that make working on your mobile phone or tablet easier and more efficient. Check out some of what’s new below: Notification settings in Google Docs, Sheets, and Slides Android apps As of last November, you can receive mobile notifications when a Drive file is shared with you or someone requests access to a Drive file you own. Previously, the only way to disable those notifications was via a setting in Drive. With this launch, we’ll allow you to enable or disable those notifications from the Settings menus in the Docs, Sheets, and Slides Android apps as well. View existing Named Ranges in the Google Sheets Android app In Google Sheets on the web, you can name a range of cells so that they’re easier to keep track of and find later and to simplify the process of creating formulas. With this launch, you can now view those named ranges and quickly navigate to their locations in the Google Sheets app for Android.
Launch Details Release track: Launching to both Rapid release and Scheduled release
Rollout pace: Gradual rollout (potentially longer than 3 days for feature visibility)
Tip: A large portion of #AdMobSAC16 is scored on the success of your app (such as the number of downloads and ratings – check the judging criteria for the full details) – you will need to give yourself enough time to put effort into promoting your app.
Complete final testing and make refinements based on user feedback. Your app should be high quality, and part of that is responding to user suggestions.
Release your app on an app store and start promoting it.
Today, the new Google Forms will become the default option for Form creators. This follows our launch from last September, where a revamped Google Forms was released as opt-in-only until we could add additional features and functionality. Once launched, Form creators will be redirected to the new Google Forms whenever they create a new Google Form.
To opt-out, users may click the “running man” icon in the bottom left-hand corner.
View individual responses from the editor In addition to the summary of responses currently available in the Forms editor, creators will now be able to see individual survey responses as well. This will save valuable time when analyzing Forms. At the same time, if there’s so many responses that there may be performance issues, Google Forms will let users know how to visit the responses separately, instead of slowing down the editor.
Get notified for every form response Form creators will now be able to configure Google Forms to send them email notifications whenever someone responds. The default is set to off, but users can enable it from the form’s settings menu.
Track responses to your Google Form Form creators will be able to see who they’ve sent their forms to and who still needs to respond. There will also be a new option to send out follow-up reminders as needed.
Responders will only see the new Google Forms for those forms created or edited in it.
To opt out and go back to the old Google Forms, simply click the running man icon in the bottom left-hand corner of the Google Form and follow the prompts.
Launch Details Release track: Form templates: Launching to both Rapid release and Scheduled release on Feb 10, 2016 Most features: Launching to Rapid release on Feb 10, 2016, with Scheduled release coming in 2 weeks
Rollout pace: Full rollout (1-3 days for feature visibility)
We’re delighted to announce the availability of the People API. With it, you can retrieve data about an authenticated user’s connections from their Contacts. Previously, developers had to make multiple calls to the Google+ API for user profiles and the Contacts API for contacts. The new People API uses the newest protocols and technologies and will eventually replace the Contacts API which uses the GData protocol.
For example, if your user has contacts in her private contact list, a call to the API (if she provides consent to do so) will retrieve a list containing the contacts merged with any linked profiles. If the user grants the relevant scopes, the results are returned as a people.connections.list object. Each person object in this list will have a resourceName property, which can be used to get additional data about that person with a call to people.get.
The API is built on HTTP and JSON, so any standard HTTP client can send requests to it and parse the response. However, applications need to be authorized to access the APIs so you will need to create a project on the Google Developers Console in order to get the credentials you need to access the service. All the steps to do so are here. If you’re new to the Google APIs and/or the Developers Console, check out this first in a series of videos to help you get up-to-speed.
In addition to merging data from multiple sources and APIs into a single cohesive data source, the new People API also exposes additional data that was not possible to get before, such as private addresses, phone numbers, e-mails, and birthdays for a user who has given permission.
We hope that these new features and data along with simplified access to existing data inspires you to create the next generation of cool web and mobile apps that delight your users and those in their circles of influence. To learn more about the People API, check out the official documentation here.
What is Google Data?
Google Data is the only site where you can get news from 60+ official Google blogs all in one place.
We have published 23,218 official posts since January 2005.