Monthly Archives: November 2017

Fact-checking the French election: lessons from CrossCheck, a collaborative effort to combat misinformation

Nine months ago, 37 newsrooms worked together to combat misinformation in the run-up to the French Presidential election. Organized by First Draft, and supported by the Google News Lab, CrossCheck launched a virtual newsroom, where fact-checkers collaborated to verify disputed online content and share fact-checked information back to the public.


The initiative was a part of the News Lab’s broader effort to help journalists curb the spread of misinformation during important cultural and political moments. With a recent study finding that nearly 25% of all news stories about the French Presidential election shared on social media were fake, it was important for French newsrooms to work closely together to combat misinformation in a timely fashion. 


Yesterday at our office in Paris, alongside many of the newsrooms who took part in the initiative, we released a report on the project produced by academics from the University of Toulouse and Grenoble Alpes University. The report explored the impact the project had on the newsrooms and journalists involved, and the general public.

  A few themes emerged from the report:

  • Accuracy in reporting rises above competition. While news organizations operate in a highly competitive landscape, there was broad agreement that “debunking work should not be competitive” and should be “considered a public service." That spirit was echoed by the willingness of 100 journalists to work together and share information for ten weeks leading up to Election Day. Many of the journalists talked about the sense of pride they felt doing this work together. As one journalist put it, “debunking fake news is not a scoop.”    
  • The initiative helped spread best practices around verification for journalists. Journalists interviewed for the report discussed the value of the news skills the picked up around fact-checking, image verification, and video authentication—and the lasting impact that would have on their work. One journalist noted, “I strengthened my reflexes, I progressed in my profession, in fact-checking, and gained efficiency and speed working with user generated content.” 
  • Efforts to ensure accuracy in reporting are important for news consumers. The project resonated with many news consumers who saw the effort as independent, impartial and credible (reinforced by the number of news organizations that participated).  By the end of the election, the CrossCheck blog hit nearly 600,000 page views, had roughly 5K followers on Twitter 180K followers on Facebook (where its videos amassed 1.2M views). As one news reader noted, ““many people around me were convinced that a particular piece of misinformation was true before I demonstrated the opposite to them,” said one person. “This changed how they voted.”

You can learn more about the News Lab’s efforts to work with the news industry to increase trust and fight misinformation here.

Fact-checking the French election: lessons from CrossCheck, a collaborative effort to combat misinformation

Nine months ago, 37 newsrooms worked together to combat misinformation in the run-up to the French Presidential election. Organized by First Draft, and supported by the Google News Lab, CrossCheck launched a virtual newsroom, where fact-checkers collaborated to verify disputed online content and share fact-checked information back to the public.


The initiative was a part of the News Lab’s broader effort to help journalists curb the spread of misinformation during important cultural and political moments. With a recent study finding that nearly 25% of all news stories about the French Presidential election shared on social media were fake, it was important for French newsrooms to work closely together to combat misinformation in a timely fashion. 


Yesterday at our office in Paris, alongside many of the newsrooms who took part in the initiative, we released a report on the project produced by academics from the University of Toulouse and Grenoble Alpes University. The report explored the impact the project had on the newsrooms and journalists involved, and the general public.

  A few themes emerged from the report:

  • Accuracy in reporting rises above competition. While news organizations operate in a highly competitive landscape, there was broad agreement that “debunking work should not be competitive” and should be “considered a public service." That spirit was echoed by the willingness of 100 journalists to work together and share information for ten weeks leading up to Election Day. Many of the journalists talked about the sense of pride they felt doing this work together. As one journalist put it, “debunking fake news is not a scoop.”    
  • The initiative helped spread best practices around verification for journalists. Journalists interviewed for the report discussed the value of the news skills the picked up around fact-checking, image verification, and video authentication—and the lasting impact that would have on their work. One journalist noted, “I strengthened my reflexes, I progressed in my profession, in fact-checking, and gained efficiency and speed working with user generated content.” 
  • Efforts to ensure accuracy in reporting are important for news consumers. The project resonated with many news consumers who saw the effort as independent, impartial and credible (reinforced by the number of news organizations that participated).  By the end of the election, the CrossCheck blog hit nearly 600,000 page views, had roughly 5K followers on Twitter 180K followers on Facebook (where its videos amassed 1.2M views). As one news reader noted, ““many people around me were convinced that a particular piece of misinformation was true before I demonstrated the opposite to them,” said one person. “This changed how they voted.”

You can learn more about the News Lab’s efforts to work with the news industry to increase trust and fight misinformation here.

‘Tis the season to Fi it Forward

With the season for giving right around the corner, we’re excited to kick off the Fi it Forward referral challenge. The challenge is rolling out today starting on desktop.


Like our last referral challenge, participants will earn prizes for the referrals they make throughout the challenge. In the Fi it Forward challenge, you can win up to two hardware gifts when you refer friends to Project Fi: a Google Chromecast and the new Android One moto x4.


But we’re most excited about our opportunity to pay it forward with our third gift. At the end of the challenge, Project Fi will donate $50,000 to the Information Technology Disaster Resource Center (ITDRC). We’re thrilled to see organizations like the ITDRC harness the power of communications technology to make a meaningful difference in crisis response and recovery, and we’re grateful to come together as a community to support their initiatives. Project Fi users don’t have to take any action to participate in the community gift—you’re already supporting the ITDRC’s disaster relief efforts just by being a part of Project Fi.


Ready to get started?. Remember to enter the challenge and get your referrals in by December 17. We can’t wait to Fi it Forward with all of you this holiday season.

Lights, shadows and silhouettes by #teampixel

Shadows don’t always have to be scary—they can be downright magical. This week, #teampixel is sharing everything from a solitary lemon’s shadow to palm trees silhouetted against a vivid sky in Venice, CA. Come chase shadows with us and see what you find.

If you’d like to be featured on @google and The Keyword, tag your Pixel photos with #teampixel and you might see yourself next.

7 ways the Assistant can help you get ready for Turkey Day

Thanksgiving is just a few days away and, as always, your Google Assistant is ready to help. So while the turkey cooks and the family gathers, here are some questions to ask your Assistant. 

thanksgiving

  • Show up to dinner on time: “Ok Google, how’s traffic?”
  • Prepare accordingly: “Ok Google, set a turkey timer for 4 hours.”
  • And don’t forget dessert: “Ok Google, add apple pie and pumpkin pie to my shopping list”
  • Play a game while you wait for turkey: “Ok Google, play Thanksgiving Mad Libs” 
  • Hear a funny tale: “Ok Google, tell me a turkey story” 
  • Learn something new: “Ok Google, give me a fun fact about Thanksgiving”
  • When Thanksgiving’s over, get ready for the next occasion:  “Ok Google, play holiday music” 

Happy Thanksgiving 🦃

Developing a VR game in just two weeks

Earlier this year, 3D modeler Jarlan Perez joined the Blocks team for a two-week sprint. The goal of his time with the team was to create a fully immersive virtual reality game in just two weeks using Blocks and Unreal Engine, two tools that have significantly influenced his process as a modeler and game enthusiast.

The result was “Blocks Isle,” the first level of a game that takes you on a journey to find your long lost friend in a sci-fi land of wonder. To win, you must solve a puzzle using hidden clues and interactions throughout the experience.

Blocks Isle - Scenes.gif

You start out on a strange desert island. After uncovering some clues and pulling a handy lever, a rocky pathway opens for exploration. Up ahead, hidden radios and books reveal clues to solve the puzzle.

Getting on Blocks Isle - SMALL.gif

Initial steps to get onto Blocks Isle. Levers and teleportation immerse the user in a new world.

Blocks Isle Clip.gif
Solving the puzzle on Blocks Isle

We caught up with Jarlan to hear more about his process and advice for other developers building immersive experiences using Blocks and Unreal Engine 4.

Brittany: Tell us about using Blocks and Unreal to develop a game in such a short amount of time.

Jarlan: Tag teaming both pieces of software worked very well! Blocks allowed me to visualize and be in the space during the modeling and conceptual phase. Unreal is like giving an artist magical powers: I’m able to fully build a proof of concept and implement functionality without having to be a professional programmer.

I found myself spending part of the day in Blocks experimenting with concepts and the rest in Unreal creating basic functionality for those ideas. This method allowed for rapid prototyping and was later beneficial when populating the space with art assets.

blocks4
Basic prototype in Unreal

What tips and tricks did you uncover that made it easy to build your game?

Being able to build large parts of the environment while standing smack dab in the middle of it is wonderful.

A big thing that I found myself doing is blowing the scene up to actual size, standing in it, and using a combination of the move grip and me moving my arms back and forth to simulate walking within the space. It helped me further understand how I wanted the player to navigate the space and where certain things needed to be placed. Again all within Blocks and no code.
Blocks Isle - Simulated Walking - SMALL.gif
Simulating walking through the experience in Blocks, as part of the creation process

Another general tip, the snap trigger is your friend! I’ve used it for most of my modeling in Blocks to snap and place assets.

Blocks Isle - Snapping - SMALL.gif
Using Blocks’ snapping feature to align shapes in the environment

How did you experiment with different ideas and concepts?


I had a few different concepts when I started the project. Blocks allowed me to quickly build a mock up of each for testing.

Blocks is an amazing tool for spatial prototyping. Before bringing a scene into Unreal, I’d blow it up to scale and move around in the space to see if it makes sense for what I’m trying to achieve. This saved me so much time.
BlocksIsle
Further development of the Blocks Isle concept

Without Blocks, how might this process have been different?

After all is said and done, I still had to take the geometry from Blocks and bring it into a 3D program for unwrapping and lightmap baking.

That said, even though I am proficient in traditional 3D modeling, I think the project would have taken longer to put together without Blocks. Blocks helped me take out some steps in the process. Traditionally I’d model out the scene and export pieces as I went, bringing them into the engine, placing them, and moving around to get a sense of how the space feels. All that got combined inside Blocks. Oh, and not to mention color exploration. If I wanted to try out colors I’d also have to create materials and place them on each asset during the in-engine test which takes more time. I can easily preview all of that in Blocks.

What advice would you give to other game developers about using these tools?


Keep exploring and always stay hungry. Be on the lookout for new tools that can improve your process and don’t be afraid of trying something new. If it doesn’t work out, it’s ok. We learn so much more from the challenges we take on than from the ones we don’t face by walking the easy path.

There are some amazing low poly games and artists out there. I think many artists would benefit from making models in VR using Blocks. If I was able to finish this project in two weeks, I can only imagine what a small team could do. Give it a try, and post your creations or questions using #MadeWithBlocks.

If you’d like to experience Blocks Isle on the HTC Vive, you can download the game.

The High Five: our searches go on, and on

Turkey, “Titanic” and the pope’s new ride were on our minds this week. Here are a few of the week’s top search trends, with data from the Google News Lab.

Almost time for turkey

As people in the U.S. prepare to gather around the table for Thanksgiving next week, our Thanksgiving insights page has all the trends. Pumpkin pie dominates searches in the U.S., but pecan pie is more popular in the southeast and apple pie is the state favorite in New Hampshire and Massachusetts. A smoked turkey is popular in most states, though some contend it should be roasted, fried or grilled. And Friendsgiving continues to rise in popularity, with searches like “friendsgiving ideas,” “friendsgiving invitations” and “friendsgiving games.”

We’ll never let go

Two decades ago, “Titanic” left an iceberg-sized hole in our hearts, and now it’s coming back to theaters in honor of its 20-year anniversary. In the years since its debut, search interest in “Titanic” reached its highest point globally in April 2012 when Titanic in 3D was released. All this talk of sinking ships made us think about other famous boats—the top searched shipwrecks this week include the Batavia, the Edmund Fitzgerald and the USS Indianapolis.

Hot wheels

The “popemobile” got an upgrade this week. Lamborghini gifted the pope a special edition luxury car, which he decided to auction off for charity. Though the pope is known for his affinity for Fiats, interest in “Pope Lamborghini” zoomed 190 percent higher than “Pope Fiat.” People also searched to find out, “Why did the Lamborghini company give the pope a car?” and “How much does the Lamborghini that they gave the pope cost?”

That’s a foul

Searches for “UCLA basketball players” shot 330 percent higher this week when three players returned home after being arrested for shoplifting while on tour with the team in China. The search queries dribbled in: “How long are the UCLA players suspended for?” “Why did China let the UCLA players go?” and “What were the UCLA players stealing?”

All about the music

With hits like “Despacito” and “Mi Gente” taking over the globe this year, the Latin Grammys last night were a hot ticket. People searched “How to watch the Latin Grammy awards online,” “What time are the Latin Grammy awards on?” and “How does music qualify for a Latin Grammy award?” Of the nominees for Record of the Year, “Despacito,” “Guerra,” and “Felices Los 4” were the most searched.

Source: Search


The High Five: our searches go on, and on

Turkey, “Titanic” and the pope’s new ride were on our minds this week. Here are a few of the week’s top search trends, with data from the Google News Lab.

Almost time for turkey

As people in the U.S. prepare to gather around the table for Thanksgiving next week, our Thanksgiving insights page has all the trends. Pumpkin pie dominates searches in the U.S., but pecan pie is more popular in the southeast and apple pie is the state favorite in New Hampshire and Massachusetts. A smoked turkey is popular in most states, though some contend it should be roasted, fried or grilled. And Friendsgiving continues to rise in popularity, with searches like “friendsgiving ideas,” “friendsgiving invitations” and “friendsgiving games.”

We’ll never let go

Two decades ago, “Titanic” left an iceberg-sized hole in our hearts, and now it’s coming back to theaters in honor of its 20-year anniversary. In the years since its debut, search interest in “Titanic” reached its highest point globally in April 2012 when Titanic in 3D was released. All this talk of sinking ships made us think about other famous boats—the top searched shipwrecks this week include the Batavia, the Edmund Fitzgerald and the USS Indianapolis.

Hot wheels

The “popemobile” got an upgrade this week. Lamborghini gifted the pope a special edition luxury car, which he decided to auction off for charity. Though the pope is known for his affinity for Fiats, interest in “Pope Lamborghini” zoomed 190 percent higher than “Pope Fiat.” People also searched to find out, “Why did the Lamborghini company give the pope a car?” and “How much does the Lamborghini that they gave the pope cost?”

That’s a foul

Searches for “UCLA basketball players” shot 330 percent higher this week when three players returned home after being arrested for shoplifting while on tour with the team in China. The search queries dribbled in: “How long are the UCLA players suspended for?” “Why did China let the UCLA players go?” and “What were the UCLA players stealing?”

All about the music

With hits like “Despacito” and “Mi Gente” taking over the globe this year, the Latin Grammys last night were a hot ticket. People searched “How to watch the Latin Grammy awards online,” “What time are the Latin Grammy awards on?” and “How does music qualify for a Latin Grammy award?” Of the nominees for Record of the Year, “Despacito,” “Guerra,” and “Felices Los 4” were the most searched.

An AI Resident at work: Suhani Vora and her work on genomics

Suhani Vora is a bioengineer, aspiring (and self-taught) machine learning expert, SNES Super Mario World ninja, and Google AI Resident. This means that she’s part of a 12-month research training program designed to jumpstart a career in machine learning. Residents, who are paired with Google AI mentors to work on research projects according to their interests, apply machine learning to their expertise in various backgrounds—from computer science to epidemiology.

I caught up with Suhani to hear more about her work as an AI Resident, her typical day, and how AI can help transform the field of genomics.

Phing: How did you get into machine learning research?

Suhani: During graduate school, I worked on engineering CRISPR/Cas9 systems, which enable a wide range of research on genomes. And though I was working with the most efficient tools available for genome editing, I knew we could make progress even faster.

One important factor was our limited ability to predict what novel biological designs would work. Each design cycle, we were only using very small amounts of previously collected data and relied on individual interpretation of that data to make design decisions in the lab.

By failing to incorporate more powerful computational methods to make use of big data and aid in the design process, it was affecting our ability to make progress quickly. Knowing that machine learning methods would greatly accelerate the speed of scientific discovery, I decided to work on finding ways to apply machine learning to my own field of genetic engineering.

I reached out to researchers in the field, asking how best to get started. A Googler I knew suggested I take the machine learning course by Andrew Ng on Coursera (could not recommend it more highly), so I did that. I’ve never had more fun learning! I had also started auditing an ML course at MIT, and reading papers on deep learning applications to problems in genomics. Ultimately, I took the plunge and and ended up joining the Residency program after finishing grad school.  

Tell us about your role at Google, and what you’re working on right now.

I’m a cross-disciplinary deep learning researcher—I research, code, and experiment with deep learning models to explore their applicability to problems in genomics.

In the same way that we use machine learning models to predict the objects are present in an image (think: searching for your dogs in Google Photos), I research ways we can build neural networks to automatically predict the properties of a DNA sequence. This has all kinds of applications, like predicting whether a DNA mutation will cause cancer, or is benign.

What’s a typical day like for you?

On any given day, I’m writing code to process new genomics data, or creating a neural network in TensorFlow to model the data. Right now, a lot of my time is spent troubleshooting such models.

I also spend time chatting with fellow Residents, or a member of the TensorFlow team, to get their expertise on the experiments or code I’m writing. This could include a meeting with my two mentors, Mark DePristo and Quoc Le, top researchers in the field of machine learning who regularly provide invaluable guidance for developing the neural network models I’m interested in.

What do you like most about the AI Residency program? About working at Google?

I like the freedom to pursue topics of our interest, combined with the strong support network we have to get things done. Google is a really positive work environment, and I feel set up to succeed. In a different environment I wouldn’t have the chance to work with a world-class researcher in computational genomics like Mark, AND Quoc, one of the world’s leading machine learning researchers, at time same time and place. It’s pretty mind-blowing.

What kind of background do you need to work in machine learning?

We have such a wide array of backgrounds among our AI Residents! The only real common thread I see is a very strong desire to work on machine learning, or to apply machine learning to a particular problem of choice. I think having a strong background in linear algebra, statistics, computer science, and perhaps modeling makes things easier—but these skills are also now accessible to almost anyone with an interest, through MOOCs!

What kinds of problems do you think that AI can help solve for the world?

Ultimately, it really just depends how creative we are in figuring out what AI can do for us. Current deep learning methods have become state of the art for image recognition tasks, such as automatically detecting pets or scenes in images, and natural language processing, like translating from Chinese to English. I’m excited to see the next wave of applications in areas such as speech recognition, robotic handling, and medicine.

Interested in the AI Residency? Check out submission details and apply for the 2018 program on our Careers site.

Android Things Contest Winners

Posted by Dave Smith, Developer Advocate for IoT

Back in September, we worked with Hackster.io to encourage the developer community to build smart connected devices using Android Things and post their projects to the Developer Challenge for Android Things. The goal was to showcase the combination of turnkey hardware and a powerful SDK for building and maintaining devices at scale.

Thank you to everyone who participated in the contest and submitted a project or idea. We had over 1100 participants register for the contest, resulting in over 350 submissions. Out of that group, we've chosen three winners. Each winner will receive support and tools from Dragon Innovation to develop their concepts into commercial products. Join us in congratulating the following makers!

Best Enterprise Project: Distributed Air Quality Monitoring

Maker: James Puderer

Monitor air quality on a street-by-street level using Android Things, Google Cloud IoT Core, and taxis!

This project showcases how Android Things makes it easy to build devices that integrate with the various services provided by the Google Cloud Platform for robust data collection and analysis. It's a clever end-to-end solution that shows understanding of both the problem domain as well as the technology.

Best Start Up Project: BrewCentral

Maker: Trent Shumay and Steven Pridie

Brewing amazing beer is a balance of art, science, and ritual. The BrewCentral system makes it possible for anyone to do an all-grain brew!

BrewCentral pairs a real-time PID controller with the touch-enabled UI and decision-making compute power of Android Things. The result is a system that accurately controls the time, temperature, and flow rates necessary to achieve repeatable results during a brew cycle. The planned enhancements for cloud-based brewing recipes will make this a connected experience for the entire brewing community.

Best IoT Project: BrailleBox - Braille News Reader

Maker: Joe Birch

BrailleBox is a small piece of hardware that empowers users who are hard-of-sight to read the latest news articles in Braille.

This project is a great use case of using IoT to have a social impact. The current proof of concept streams articles from a news feed to the Braille pad, but this project has the potential to leverage machine learning on the device to translate additional input from the physical world into a Braille result.

Honorable Mentions

The community submitted some amazing projects for the contest, which made the choice of picking only three winners extremely difficult. Here are a few of our favorite projects that weren't selected for a prize:

  • Andro Cart: A shopping cart device powered by Android Things. Designed to help decentralize point of sale (POS) billing.
  • SIGHT: For the Blind: A pair of smart glasses for the blind, powered by Android Things and TensorFlow.
  • Industrial IoT Gateway: A smart industrial gateway for the IoT world based on Android Things.
  • Sentinel: The first semi-autonomous home security robot based on Android Things.
  • Word Clock: A creative take on reading the time, powered by Android Things. Control it via the Nearby API or the Google Assistant.

We encourage everyone to check out all the new projects in the Google Hackster community, and submit your own as well! You can also join Google's IoT Developers Community on Google+, a great resource to get updates, ask questions, and discuss ideas. We look forward to seeing what exciting projects you build!