May 2017 Summaries

Christina Hayek, an Arabic Language Manager at Google, highlights the importance of localization through a personal story about gifting her grandmother an Android smartphone that functioned in Arabic, showcasing the efforts of Google's localization team. This team, spread across more than 30 countries, ensures Google products are accessible and culturally relevant in over 70 languages by adapting content beyond mere translation, such as changing baseball to football in Italy. To address the growing demand for skilled localization professionals, Google, in collaboration with Udacity, launched a free course called Localization Essentials. This massive open online course (MOOC) aims to equip learners worldwide with the foundational knowledge needed to create globally accessible products, particularly as the internet user base expands rapidly in non-English speaking regions. The course represents a significant collaborative effort and seeks to provide educational opportunities in a field that is crucial for developing culturally diverse and relevant user experiences.

At Google I/O 2017, Firebase announced several enhancements aimed at improving app development and user engagement. The platform integrated with Fabric to incorporate Crashlytics as its primary crash reporting tool and introduced phone number authentication, broadening its authentication solutions. Firebase Performance Monitoring was launched in beta to help developers address app performance issues, while Firebase Analytics was rebranded as Google Analytics for Firebase to reflect its collaboration with Google Analytics. The updates also included support for iOS and game developers, open-sourcing SDKs, and integrating Cloud Functions with Firebase Hosting for dynamic content delivery. These advancements underscore Firebase's commitment to providing a robust, developer-friendly environment, further supported by the introduction of a Firebase Alpha program for community-driven innovation.

Women Techmakers, a global program by Google aimed at supporting women in technology, announced a new scholarship in partnership with Udacity to help women worldwide enhance their technical skills through online degrees. This initiative follows the success of increased female participation at Google I/O, where 25% of attendees and speakers were women, highlighting the growing influence of women in tech. The Women Techmakers Udacity Scholarship offers opportunities for women at various skill levels to pursue certifications in Android Basics, Android Developer, Front-End Web Developer, and Full-Stack Developer, with courses co-designed by Google experts. Recipients will gain access to a supportive community of Googlers and peers worldwide, and upon completing the Nanodegree within a year, they will receive certification from Udacity, a certificate from Women Techmakers, and a resume review by a Googler. This scholarship is part of Women Techmakers' broader efforts to provide resources, community, and visibility to women and allies in tech, complementing other initiatives such as personalized professional support, video series, guides for building inclusive tech hubs, and a Scholars program for university women in computer science.

Launchpad Accelerator's fourth class was announced on May 23, 2017, featuring a diverse group of startups from Asia, Latin America, Africa, and Europe. The program, beginning on July 17, 2017, at the Google Developers Launchpad Space in San Francisco, aims to assist these startups in leveraging Google's machine learning and AI technologies. Participants will benefit from intensive mentoring from over 20 Google teams and experts from Silicon Valley's top tech companies and venture capitalists, receiving equity-free support and Google product credits. The accelerator offers a six-month program that includes two weeks of all-expenses-paid training, followed by continued collaboration with Google in their home countries. The participating startups span a wide range of sectors, including financial services, digital commerce, HR solutions, food and retail, education, and health, aiming to scale their operations and innovate within their respective industries.

At the 2017 Google I/O conference, Google introduced several new tools and updates aimed at enhancing user experiences and aiding developers in growing their businesses. Key announcements included the expansion of Google's payment solutions, allowing users to easily pay with saved credit and debit cards through the Google Payment API, which promises faster checkouts and increased sales conversion for developers. Additionally, a redesigned AdMob platform was unveiled, integrating Google Analytics for Firebase to provide deeper insights into user behavior and ad revenue. Google also enhanced its Universal App Campaigns (UAC) to optimize app promotion through machine learning and introduced new ad placements in the Google Play Store to connect users with apps during discovery moments. Smart Bidding strategies in UAC were expanded to target high-value users, and a new App Attribution Partners program was launched to integrate third-party measurement data directly into AdWords, providing more detailed app engagement metrics. These innovations reflect Google's commitment to creating user-centric experiences and supporting developers in a rapidly evolving digital landscape.

The state of the modern mobile web has significantly evolved, as highlighted during Google's I/O event, with technologies like Accelerated Mobile Pages (AMP) and Progressive Web Apps (PWAs) gaining widespread adoption. These innovations have enabled businesses to enhance mobile site performance, decrease load times, and improve user engagement, as seen with companies like Wego, Forbes, Ola, and Twitter, which have reported substantial increases in conversion rates and user interaction through the integration of PWAs. Developers have been provided with new tools and APIs, such as the Improved Add to Homescreen, Payment Request API, and Media Session API, which enhance mobile web capabilities by enabling deeper integration with Android systems, streamlining payment processes, and improving media consumption experiences. Additionally, tools like Lighthouse and Polymer 2.0 have been developed to support the creation of high-quality web experiences, emphasizing Chrome's commitment to fostering a thriving web ecosystem.

Google I/O 2017, held at the Shoreline Amphitheatre, brought together 7,200 attendees and millions of online viewers to showcase Google's efforts in simplifying the development process across its billion-user platforms like Android and Chrome. The event highlighted significant updates and new tools such as the Android O Developer Preview 2, Project Treble for faster Android updates, and Kotlin support for Android development to enhance developer experience and app performance. Google emphasized the importance of optimizing mobile experiences with Android Go, AMP, and Progressive Web Apps, and introduced features in Firebase to improve app development and performance monitoring. New certifications for web developers and enhancements in the Google Play Console, including the release of Android Instant Apps, were also announced to help developers grow their businesses. The introduction of a Google Payment API and integration of AdMob with Firebase sought to streamline monetization, while updates to Actions on Google for the Google Assistant and developments in VR and AR platforms underscored Google's commitment to innovation. The conference also reinforced Google's dedication to community growth through global initiatives and training programs, with future events planned in Europe and India.

The TensorFlow Research Cloud (TFRC) is an initiative launched by Google to provide researchers with free access to a cluster of 1,000 Cloud TPUs, designed to support computationally-intensive machine learning projects that might otherwise be unattainable. With up to 180 teraflops of performance per TPU and 64 GB of high-bandwidth memory, the TFRC aims to accelerate open machine learning research by enabling the development of new, more accurate models. The program is open to a diverse range of applicants, not limited to academia, and encourages contributions from individuals with non-traditional backgrounds. Successful applicants are expected to share their research publicly and provide feedback to improve the program. Additionally, Google offers a Cloud TPU Alpha program for businesses seeking to accelerate proprietary model training and data processing. The overarching goal of the TFRC is to advance the machine learning research community by fostering innovative discoveries.

Google is enhancing its Assistant platform for developers by extending Actions on Google to Android phones and iPhones, allowing for new use cases and UI elements like image carousels and suggestion chips. The update includes the introduction of transaction and payment capabilities, offering two options: Google-facilitated payments or user-provided payment methods with seamless account linking. A new developer console integrated with Firebase and Google Cloud helps teams track app performance and user discovery, while an app directory improves app discoverability with categories and user ratings. The Assistant SDK has been updated to support hotwords and alarms, and a developer competition has been announced to encourage the creation of innovative apps for Google Assistant. The platform is also set to expand to new languages and devices, reflecting ongoing efforts to enhance the developer experience.

Material Components offers a set of open-source UI components for building applications and websites across Android, iOS, and web platforms, leveraging Material Design principles to ensure cohesive style, brand, interaction, and motion. Developed and maintained by a team of Google engineers and designers, these components are frequently updated to conform to the latest Material Design guidelines, embracing internationalization and accessibility standards, and ensuring pixel-perfect accuracy. The components support older Android versions through graceful degradation, adhere to industry standards in iOS development with Swift and Objective-C, and integrate seamlessly with popular web frameworks and libraries. By using these components, developers can create rich user experiences while contributing to the ongoing evolution of the project, with resources available on GitHub for collaboration and updates.

Google I/O 2017, hosted at the Shoreline Amphitheatre from May 17-19, promises a schedule full of technical content, product demos, and interactive experiences for attendees. Participants can reserve seats for sessions via web and mobile apps and engage in various activities such as Sandbox demos, Codelabs, and Office Hours to interact with Google engineers. Evening events include an After Hours Block Party and a concert, featuring entertainment such as a comedy club, international food market, and a Museum of Developer Art. For those attending remotely, options include watching the livestream, participating in I/O Extended events, and engaging through social media using #io17request. Attendees are encouraged to dress appropriately for the outdoor venue and can access logistical information about transportation and parking.

In response to a phishing attack exploiting the OAuth authorization infrastructure, Google announced updates to developer identity guidelines and registration processes to enhance user protection. These updates involve stricter app identity guidelines, requiring app names to be unique and not misleading, and improvements to the app publishing process, risk assessment systems, and user consent pages to detect spoofed identities. A new manual review process has been introduced for web applications requesting user data, which may delay the ability to publish apps but aims to prevent unauthorized access. Developers are encouraged to plan accordingly and can still test apps using registered accounts before approval. Additionally, Google emphasizes the importance of developers understanding their responsibilities when accessing user data to maintain a secure developer ecosystem.

In May 2017, the TensorFlow team released a collection of performance benchmarks demonstrating the speed and scalability of TensorFlow when training image classification models such as InceptionV3 and ResNet on various hardware configurations. These benchmarks, including tests on NVIDIA's DGX-1 and Tesla K80 GPUs in both single-server and distributed setups, showed near-linear performance scaling and significant speedups, particularly when using synthetic data. The team also introduced a High-Performance Models guide to assist developers in optimizing their models for different platforms. The benchmarks highlighted the importance of tailoring model configurations to specific hardware setups, as optimal performance varied based on the platform, with some models performing better when variables were handled differently across GPUs. The team emphasized the need for comprehensive benchmarking to achieve optimal performance and discussed plans for future tests focusing on convergence time to high accuracy levels. They acknowledged NVIDIA's support in providing hardware and technical assistance, expressing anticipation for future collaborations to enhance TensorFlow's performance, especially with NVIDIA's forthcoming Volta architecture.

Google I/O 2017 is approaching, and attendees can customize their experience using the Google I/O 2017 mobile app, which is available for Android and iOS devices, as well as a web app. The app allows users to browse, filter, and schedule sessions, office hours, and codelabs across 14 tracks, with all sessions being livestreamed and recorded for later access. Attendees can sync their schedules across devices and receive reminders for saved sessions. Onsite registered attendees can reserve seats for sessions in advance, with reservations being managed through the app and the Google I/O website. The app also provides a custom conference map, event announcements, and additional information such as onsite WiFi details and travel tips. The app will be open-sourced after the event, and users are encouraged to utilize it for a seamless experience at the conference.

AIY Projects, launched by Google, provides do-it-yourself artificial intelligence resources for Makers, enabling them to integrate AI into their projects and make human-to-machine interactions resemble human-to-human interactions. The initiative begins with the Voice Kit, a project that allows Makers to build a Voice User Interface using cloud services like the Google Assistant SDK or the Cloud Speech API, or to operate entirely on-device, extending the capabilities of Raspberry Pi. The Voice Kit includes hardware for audio capture and playback, connectors for additional components, and offers compatibility with Android Things for prototyping IoT devices. The kit is available through various retailers and includes comprehensive instructions and resources online, encouraging community engagement and feedback through platforms like hackster.io and social media.