Google Keynote (Google I/O ‘23)

Introduction

The speaker talks about how they set out to improve people's lives and how AI will help them achieve that goal.

• Google aims to significantly improve the lives of as many people as possible.

• They have learned a lot from their users, including discovering new places and learning new words.

• With breakthroughs in AI, they aim to reimagine the ways they can help people and businesses.

• Their approach will always be responsible because their goal is to make AI helpful for everyone.

Opening Remarks

Sundar Pichai welcomes everyone to Google I/O and talks about how AI is having a busy year.

• Sundar Pichai welcomes everyone to Google I/O.

• He thanks developers for attending and those joining from around the world.

• He mentions that AI is having a busy year, and there's lots to talk about.

Making AI More Helpful

Sundar Pichai talks about how Google has been applying AI to make their products more helpful.

• Google has been applying AI to make their products more helpful for a while now.

• With generative AI, they are taking the next step in reimagining all their core products, including Search.

• Smart Reply was launched in 2017, followed by Smart Compose which offered writing suggestions as you type.

• Help Me Write is the next step in Gmail with a much more powerful generative model. It allows users to create full drafts of emails by typing prompts of what they want.

Immersive View for Routes

Sundar Pichai introduces Immersive View for Routes on Google Maps.

• Immersive View uses AI technology to create high fidelity representations of places so users can experience them before they visit.

• Immersive View for Routes allows users to see their whole trip in advance, whether they're walking, cycling, or driving.

• Users can check today's air quality and traffic and weather conditions that might change over the next few hours.

• Immersive View for Routes will begin rolling out over the summer and launch in 15 cities by the end of the year.

Google Photos

Sundar Pichai talks about how AI has made Google Photos better.

• Google Photos was introduced at I/O in 2015 as one of their first AI-native products.

• Breakthroughs in machine learning made it possible to search photos for things like people, sunsets, or waterfalls.

• With AI technology, Google aims to do more than just search photos.

Magic Editor

In this section, the speaker introduces a new experience called Magic Editor that uses semantic understanding and generative AI to improve photos.

Features of Magic Editor

• The user can remove unwanted objects from a photo.

• The user can adjust the lighting in a photo.

• The user can adjust the pose of people in a photo.

• Magic Editor automatically recreates parts of an image that were not captured in the original shot.

AI Across Google Products

In this section, the speaker discusses how AI is being used across various Google products.

Examples of AI in Google Products

• Gmail, Photos, and Maps are just a few examples of how AI is helping users in moments that matter.

• Improving knowledge and learning, boosting creativity and productivity, enabling developers and businesses to build their own transformative products and services, and building and deploying AI responsibly are four important ways that AI is being used to advance Google's mission.

Advancing Foundation Models with PaLM 2

In this section, the speaker discusses PaLM 2, which is one of Google's fundamental research models for advancing its mission.

Capabilities of PaLM 2

• PaLM 2 is highly capable at a wide range of tasks and easy to deploy.

• PaLM 2 models deliver excellent foundational capabilities across a wide range of sizes.

• Gecko is so lightweight that it can work on mobile devices even when offline.

• Fine-tuning PaLM 2 on domain-specific knowledge can greatly improve its capabilities, as seen with Sec-PaLM and Med-PaLM 2.

• Med-PaLM 2 achieved a 9x reduction in inaccurate reasoning when compared to the model and is currently the state of the art.

Availability of PaLM 2

• PaLM 2 is now available in Preview.

Introduction

Sundar Pichai introduces the concept of inflection point and how Google DeepMind is focused on building more capable systems safely and responsibly. He also talks about Gemini, a next-generation foundation model that is still in training.

• AlphaGo, transformers, sequence-to-sequence models are some of the technologies that set the stage for the inflection point we are at today.

• Google DeepMind is focused on building more capable systems safely and responsibly using computational resources of Google.

• Gemini is a next-generation foundation model that is being trained to be multi-modal, highly efficient at tool in API integrations, and built to enable future innovations like memory and planning.

AI Responsibility

Sundar Pichai talks about AI responsibility and how Google is investing in identifying synthetically generated content whenever encountered. He explains watermarking and metadata as two important approaches to identify such content.

• As models get better and more capable, it becomes important to invest in AI responsibility.

• Watermarking embeds information directly into content in ways that are maintained even through modest image editing.

• Metadata allows content creators to associate additional contexts with original files giving you more information whenever you encounter an image.

Bard: Conversational AI Experiment

Sundar Pichai introduces Bard as an experiment for conversational AI. Sissie Hsiao shares updates on Bard's capabilities after integrating PaLM 2.

• Bard was launched as a limited-access experiment on a lightweight large language model to get feedback from users.

• With PaLM 2 integration, Bard's math, logic, and reasoning skills made a huge leap forward underpinning its ability to help developers with programming.

• Developers can bring code from Bard into their workflow like Colab or export it with Replit.

• Bard provides code citations and starting next week, it will make code citations even more precise.

Tools for Bard

Sissie Hsiao announces that tools are coming to Bard to fuel curiosity and imagination.

• Tools are coming to Bard to bring more capabilities and fuel curiosity and imagination.

Bard: The Future of Conversational AI

In this section, the speaker introduces Bard, a conversational AI that will become more visual in its responses and prompts.

Bard Becomes More Visual

• Bard will use Google Search and the knowledge graph to find relevant images.

• Images will be used to give users a better sense of what they are exploring.

• Google Lens will be integrated into Bard to allow users to prompt with images and create funny captions.

Using Bard for College Search

• Users can ask for help finding colleges with specific programs using natural language.

• Google Maps is integrated into Bard to visualize where schools are located.

• Users can organize search results as tables and add columns showing whether schools are public or private.

Expanding Capabilities

• Extensions from partners like Instacart, Indeed, Khan Academy, and many more will be added in the future.

• Adobe Firefly integration allows users to generate new images from their imagination within Bard.

Availability

• Bard is now available in over 180 countries and territories.

• Japanese and Korean languages have been added, with plans for 40 languages soon.

AI in Workspace: Next Generation Productivity Features

In this section, the speaker discusses how AI can act as a coach, thought partner, source of inspiration, and productivity booster across all apps in Workspace.

Collaborating with AI

• No bullet points provided.

Expanding Helpful Features Across More Surfaces

In this section, the speaker talks about how they have been expanding helpful features across more surfaces. They show examples of how Docs, Sheets, and Slides can help users get organized and creative with just a few words as prompts.

Docs: Job Descriptions

• Docs can help businesses create job descriptions quickly and easily.

• Users can enter a few words as a prompt for the job description.

• The model generates a starting point that users can customize to fit their needs.

• This feature saves time and effort for businesses.

Sheets: Dog Walking Business

• Sheets can help users organize information for their business.

• Users can type in a prompt like "client and pet roster for a dog walking business with rates" to create a table.

• The model generates a table with relevant columns like client info, notes, etc.

• This feature makes it easy for users to get started and stay organized.

Slides: Image Generation

• Slides can be used for storytelling at work or in personal life.

• Users can use prompts to generate images that match the content of their slides.

• The text-to-image models generate six to eight images so that users have options to choose from.

• This feature allows users to be creative and add pizzazz to their presentations.

Collaboration with AI

In this section, the speaker talks about how AI can proactively offer prompts that are contextual and change based on what the user is working on. They demonstrate this through a live demo where they use Sidekick, an AI-powered suggestion tool, to help them write a spooky story.

Sidekick Demo: Spooky Story Writing

• Sidekick is an AI-powered suggestion tool that offers contextual prompts based on what the user is working on.

• The speaker and their niece are writing a spooky story together and use Sidekick to get unstuck.

• Sidekick reads and processes the document and offers suggestions like "what happened to the golden seashell?" or "what are common mystery plot twists?"

• Users can select a suggestion, and Sidekick generates content that users can insert into their document.

• This feature allows users to collaborate with AI in real-time, making it easier for them to write creatively.

Contextual Collaboration

In this section, Aparna Chennapragada demonstrates how Google's new contextual collaboration can boost productivity and creativity. She provides examples of how Sidekick can help with tasks such as finding information for a potluck or creating speaker notes for a presentation.

Examples of Contextual Collaboration

• Aparna gives an example of using Sidekick to find information about what dishes people are bringing to a potluck.

• Sidekick finds the relevant Google Sheet and cites the source in the "Found In" section.

• The information can be inserted directly into an email.

• Another example is given where someone needs to create speaker notes for a presentation but has not prepared them yet.

• One of the suggestions from Sidekick is to create speaker notes for each slide.

• The presentation and other relevant context are sent to the model to help create these notes.

Labs Preview

Sundar Pichai introduces Labs, which allows users to preview some of the experiences across Workspace and other products.

Introducing Labs

• Sundar Pichai introduces Labs, which allows users to preview some of the experiences across Workspace and other products.

• Users can check it out at google.com/labs.

Generative AI in Search

Cathy Edwards discusses how generative AI is being used in Search. She demonstrates how Search can break down complex questions into smaller ones and provide helpful insights through an AI-powered snapshot.

Bringing Generative AI into Search

• Cathy Edwards demonstrates how Search can break down complex questions into smaller ones and provide helpful insights through an AI-powered snapshot.

• The AI-powered snapshot quickly gives users the lay of the land on a topic.

• Links are included in the snapshot to allow users to dig deeper and explore the richness of the topic.

Using Search for Research

• Cathy Edwards provides an example of using Search to research a good bike for a five-mile commute with hills.

• The AI-powered snapshot includes important considerations such as motor and battery for taking on those hills and suspension for a comfortable ride.

Google Search: Supercharged with Generative AI

In this section, the speaker talks about how Google is using generative AI to improve search results and create a more conversational experience for users.

New Features in Search

• Google is now graphing every hour to provide fresh and relevant results.

• Ads are clearly labeled and can be helpful for commercial queries.

• Users can ask follow-up questions or select suggested next steps to enter into a brand-new conversational mode.

• Conversational mode allows users to refine their search without starting over and see other search results seamlessly.

• Users can click on retailers that have products in stock, check out, and turn their attention to what really matters.

New Category of Experiences on Search

• Generative AI capabilities unlock a whole new category of experiences on Search.

• These experiences could help users create clever names for cycling clubs, craft perfect social posts, or test knowledge on bicycle hand signals.

Live Demo

• The speaker demonstrates how generative AI works by searching for information about whales with her daughter.

• She asks if she can see whales in California and gets a snapshot of insights from across the web.

• She follows up with "plush ones for kids under $40" and gets organized information from LLMs.

Feedback and Availability

• The new Search generative experience will be available in Labs along with some other experiments.

• Users in the US can join the waitlist today by tapping the Labs icon in the latest version of the Google app or Chrome desktop.

• This new experience reflects the beginning of a new chapter for Search, supercharged with generative AI.

Future of AI in Search

• The speaker shares an early look at what's to come for AI in Search.

• A video shows people asking questions like "Is a hot dog a sandwich?" and getting different answers.

• The speaker concludes by saying that Google is excited to keep evolving search and building it with users.

Google Cloud AI Platform

In this section, Thomas Kurian talks about how Google is using the power of AI to transform the way businesses work. He explains how Google Cloud can help businesses take advantage of the massive opportunity in front of them by building generative applications using their AI platform Vertex AI.

Transforming the Way You Work

• All investments made by Google are also coming to businesses.

• Thousands of companies are already using generative AI platform to create amazing content, synthesize and organize information, automate processes, and build incredible customer experiences.

• Three ways Google Cloud can help you take advantage of the massive opportunity in front of you:

• Build generative applications using our AI platform Vertex AI

• Retrieve fresh and factual information from your company's databases, corporate intranet, website, and enterprise applications with Enterprise Search

• Incorporate alternative AI technologies into business processes

Building Generative Applications with Vertex AI

• With Vertex, you can access foundation models for chat, text, and image.

• Both Vertex and Enterprise Search guarantee privacy of data.

• Choose the best model for your specific needs across many sizes that have been optimized for cost, latency, and quality.

Partnering with Other Companies

• Working with partners like character.ai to provide consumers with deeply personalized characters.

• Partnering with Salesforce to integrate Google Cloud's AI models and BigQuery with their data cloud in Einstein.

• Working with many other incredible partners including consultancies, software-as-a-service leaders, consumer internet companies, and many more to build remarkable experiences with AI technologies.

New Models in Vertex

• Introducing three new models in Vertex: Imagen, Codey, and Chirp.

• Reinforcement learning from human feedback into Vertex AI.

Duet AI for Google Cloud

In this section, Thomas Kurian introduces Duet AI for Google Cloud, which is an always-on AI collaborator that uses generative AI to provide developers assistance within the IDE, the Cloud Console, or directly within chat.

Introducing Duet AI for Google Cloud

• Duet AI for Google Cloud is an always-on AI collaborator that helps developers get things done.

• It uses generative AI to provide developers assistance wherever they need it, within the IDE, the Cloud Console, or directly within chat.

• It can provide contextual code completion and offer suggestions tuned to your code base.

• It can generate entire functions in real-time and assist with code reviews and inspection.

Building on an AI-optimized infrastructure

In this section, Thomas Kurian talks about building all of these capabilities on their AI-optimized infrastructure. This infrastructure makes large-scale training workloads up to 80% faster and up to 50% cheaper compared to any alternatives out there.

Building on an optimized infrastructure

• All of these capabilities are built on their AI-optimized infrastructure.

• The A3 Virtual Machines have been added to this infrastructure family based on Nvidia's latest H100 GPUs.

• They provide a wide choice of compute options for leading AI companies like Entropik and Midjourney.

PaLM API

In this section, Josh Woodward talks about PaLM 2 - their most capable language model that powers the PaLM API. They have been running a private preview with their PaLM API since March, and it has been amazing to see how quickly developers have used it in their applications.

PaLM API

• PaLM 2 is Google's most capable language model that powers the PaLM API.

• Developers want choice when it comes to language models, and they are working with leading developer tools companies like LangChain, Chroma, and many more to support the PaLM API.

• The PaLM API has been integrated into Google Developer tools like Firebase and Colab.

Project Tailwind

In this section, Josh Woodward introduces Project Tailwind - an AI-first notebook that helps you learn faster. It creates a personalized and private AI model that has expertise in the information that you give it.

Project Tailwind

• Project Tailwind is an AI-first notebook that helps you learn faster.

• It creates a personalized and private AI model that has expertise in the information that you give it.

• It was developed with authors like Stephen Johnson and tested at universities like Arizona State and the University of Oklahoma.

• It can pull out key concepts and questions grounded in the materials given to it.

Project Tailwind

In this section, the speaker talks about project Tailwind and how it can be helpful for anyone synthesizing information from many different sources.

Tailwind is not just for students

• Project Tailwind is not just for students.

• It's helpful for anyone synthesizing information from many different sources.

• The speaker invites people to try out the project.

PaLM 2 and Generative AI

In this section, the speaker talks about PaLM 2 and generative AI.

• There's a lot more that can be done with PaLM 2.

• Generative AI is changing what it means to develop new products.

• Google offers the best ML infrastructure, powerful models including those in Vertex, APIs, and tools to quickly generate your own applications.

Responsible Approach to Bold AI

In this section, the speaker talks about taking a responsible approach to bold AI.

• Building bold AI requires a responsible approach.

• Our field-defining research is helping scientists make bold advances in many scientific fields including medical breakthroughs.

• While there's a natural tension between being bold and responsible, we believe it's critical to embrace that tension productively.

Misinformation and Evaluating Images

In this section, the speaker talks about misinformation and evaluating images.

• Misinformation is an area of concern when it comes to generative AI.

• Google is developing tools to evaluate online information.

• Two new ways for people to evaluate images are being added, including About this Image tool in Google Search and using Google Lens.

Universal Translator

In this section, the speaker talks about Universal Translator.

• Universal Translator is an experimental video dubbing service that helps experts translate a speaker's voice while also matching their lip movements.

AI and Responsibility

In this section, the speaker talks about how Google is using next-generation translation models to translate what the speaker is saying. They also discuss how they are replicating the style and tone of speakers and matching their lip movements. The speaker then goes on to talk about the inherent tension between the benefits of this technology and its potential misuse by bad actors.

Next-Generation Translation Models

• Google uses next-generation translation models to translate what speakers are saying.

• They replicate the style and tone of speakers and match their lip movements.

Benefits vs Misuse

• There is an inherent tension between the benefits of this technology and its potential misuse by bad actors.

• Google has built guardrails into their service to help prevent misuse and make it accessible only to authorized partners.

AI Principles at Google

In this section, the speaker discusses how Google's AI principles guide them on what not to do with their technology. They also talk about innovations that tackle challenges as they emerge, such as reducing problematic outputs generated by models.

Guiding Principles

• Google's AI principles guide them on what not to do with their technology. For example, years ago, they were the first major company to decide not to make a general-purpose facial recognition API commercially available due to inadequate safeguards in place.

Innovations for Tackling Challenges

• Google develops innovations that tackle challenges as they emerge, such as reducing problematic outputs generated by models.

• They use automated adversarial testing using large language model technology for queries like these to help uncover and reduce inaccurate outputs, significantly improving speed, quality, and coverage of testing.

• Google shares these innovations with others, such as their perspective API originally created to help publishers mitigate toxicity, which is now being used in large language models.

Building AI Responsibly

In this section, the speaker emphasizes that building AI responsibly must be a collective effort involving researchers, social scientists, industry experts, governments, and everyday people. They also discuss working with the web community on ways to give publishers choice and control over their web content.

Collective Effort

• Building AI responsibly must be a collective effort involving researchers, social scientists, industry experts, governments, and everyday people.

Giving Publishers Choice and Control

• Google is working with the web community on ways to give publishers choice and control over their web content.

Bringing Computing to Everyone

In this section, the speaker talks about how Android's open OS enables a whole ecosystem and brings smartphones to everyone. They also discuss using AI to improve experiences at scale by protecting users from spam messages and calls.

Android's Open OS

• Android's open OS enables a whole ecosystem and brings smartphones to everyone.

Using AI for Improving Experiences at Scale

• Android uses AI models to protect users from more than 100 billion suspected spam messages and calls.

Android's Ecosystem of Devices

In this section, the speaker discusses how Android is bringing computing benefits to everyone by continuing to connect them to the most complete ecosystem of devices where everything works better together.

Large Screens Experience Redesign

• Over the last two years, Android has redesigned the experience on large screens, including tablets and foldables.

Optimizing Google Apps

• Google has made huge investments to optimize more than 50 Google Apps, including Gmail, Photos, and Meet.

Working with Partners

• Google is working closely with partners such as Minecraft, Spotify, and Disney Plus to build beautiful experiences that feel intuitive on larger screens.

Wear OS and Android Updates

In this section, Sameer Samat talks about the momentum in smartwatches and how Wear OS is now the fastest-growing watch platform. He also shares that WhatsApp is bringing their first-ever watch app to Wear this summer. Additionally, he discusses the importance of having all devices work seamlessly together and introduces updates to Fast Pair, Nearby Share, Cast, and Find My Device experience.

Momentum in Smartwatches

• Wear OS is now the fastest-growing watch platform.

• WhatsApp is bringing their first-ever watch app to Wear this summer.

Seamless Device Integration

• Fast Pair lets you easily connect more than 300 headphones.

• Nearby Share makes it easy to move files between your phone, tablet or Windows and Chrome OS computer.

• Cast makes streaming video and audio to your devices ultra-simple with support from over 3,000 apps.

• A major update to Find My Device experience supports a wide range of devices in your life including headphones, tablets, etc.

Privacy Protection

• The network for Find My Device experience is designed in a privacy-preserving way where location information is encrypted.

• Unknown tracker alerts will tell you if an unrecognized tracking tag is moving with you and help you locate it.

Connecting People

• RCS technology upgrades old SMS and MMS technology to make sending high-quality images and video possible across different types of phones.

Generative AI for Personalization

Dave Burke talks about how customization has always been at the core of the Android experience. This year, they are combining Android's guided customization with Google's advances in generative AI so that your phone can feel even more personal. He introduces Magic Compose, a new feature coming to Google messages powered by generative AI that helps you add an extra spark of personality to your conversation.

Personalization

• Customization has always been at the core of the Android experience.

• Magic Compose is a new feature coming to Google messages powered by generative AI that helps you add an extra spark of personality to your conversation.

Customizable Lockscreen Shortcuts and Emoji Wallpapers

In this section, the speaker talks about the new customizable lock screen shortcuts and emoji wallpapers that are available on Android devices.

Customizable Lockscreen Shortcuts

• The new customizable lock screen shortcuts allow users to instantly jump into their most frequent activities.

• Users can customize their lock screen and home screen with their favorite wallpaper.

• The speaker demonstrates how to use the new option for emojis in the wallpaper picker to create a personalized wallpaper with emojis.

• Users can choose from different patterns and colors to bring their personalized wallpaper together.

Emoji Wallpapers

• Emojis on wallpapers are a fun way of expressing oneself.

• Users can create a beautiful 3D image from any regular photo using the cinematic wallpaper feature.

• Generative AI wallpapers allow users to choose what inspires them, and then Google creates a beautiful wallpaper to fit their vision.

• Pixel has been leading in AI-driven hardware experiences for years.

Pixel: An AI-first Mobile Computer

In this section, the speaker talks about how Pixel was conceived as an AI-first mobile computer, bringing together all the amazing breakthroughs across Google into one device.

Pixel's Innovation in AI-driven Hardware Experiences

• Pixel has been leading in AI-driven hardware experiences for years.

• Other phones have AI features, but Pixel is the only phone that brings together all of Google's amazing breakthroughs into one device.

Pixel AI and Camera

In this section, the speaker talks about how Pixel's personal AI helps users save time and get more done. They also discuss the unique features of Pixel Call Assist, Pixel Speech experiences, and Pixel Safe. The speaker then goes on to talk about the unique features of the Pixel Camera.

Personal AI

• Pixel's personal AI enables helpful experiences that are not available on any other mobile device.

• On-device machine learning translates different languages for you, transcribes conversations in real-time, and understands how you talk and type.

• Pixel Safe is a collection of features that keep you safe online and in the real world.

Pixel Camera

• Pixel Camera understands faces, expressions, and skin tones to better depict you and the people you care about so your photos will always look amazing.

• Pixel camera is constantly working to make itself more inclusive with features like Real Tone and Guided Frame.

Google's Popularity Growth

In this section, the speaker talks about how Google has become one of the fastest-growing OEM in their markets. They also mention that last year's Pixel devices were their most popular generation yet with both users and respected reviewers.

Popularity Growth

• Google has become one of the fastest-growing OEM in their markets.

• Last year's Pixel devices were their most popular generation yet with both users and respected reviewers.

Introducing The New Member Of Our A Series: The Gorgeous New Pixel 7a

In this section, the speaker introduces the newest member of their a series, the Pixel 7a. They talk about how it delivers incredible Pixel performance in a very affordable device and how it has the highest-rated camera in its class.

The New Member Of Our A Series

• The Pixel 7a delivers incredible Pixel performance in a very affordable device.

• The 7a takes the crown from 6A as the highest-rated camera in its class.

• Pixel 7a is available today, starting at $499.

Introducing The Pixel Tablet

In this section, the speaker talks about how they set out to build something different with tablets and large screens. They introduce the new Pixel Tablet, which is designed specifically to be helpful in your hand and in the place they are used most - at home.

The Pixel Tablet

• Pixel Tablet is designed specifically to be helpful in your hand and in the place they are used most - at home.

• Pixel Tablet is powered by Google Tensor G2 for long-lasting battery life and cutting-edge personal AI.

• Pixel Camera on Tensor G2 is optimized specifically for video calling.

Pixel Tablet and Charging Speaker Dock

In this section, the speaker talks about the features of Pixel Tablet and how it can be helpful in the home.

Features of Pixel Tablet

• With Pixel Speech and Tensor G2, voice typing is nearly three times faster than tapping.

• The charging speaker dock gives the tablet a home and makes sure that it is always charged.

• When docked, Hub mode turns Pixel Tablet into a digital photo frame, smart home controller, voice-activated helper, and shared entertainment device.

• With an ultra-fast fingerprint sensor, users can quickly unlock the device and get immediate access to all their favorite Android apps.

Smart Home Controls

• The redesigned Google Home app brings together over 80,000 supported smart home devices including Matter-enabled devices.

• The new home panel allows any family member to quickly adjust lights or lock doors directly from Hub mode.

• Pixel Tablet supports multiple users with each user having their own apps and content while maintaining privacy.

Chromecast Built-in

• Pixel Tablet is the first tablet with Chromecast built-in making it easy to cast music or shows from phone to tablet.

• A new type of case for Pixel Tablet has a built-in stand that provides continuous flexibility at all angles so users can comfortably use their tablet anywhere.

Availability

• The new Pixel Tablet comes in three colors starting at $499 and ships next month.

• Every Pixel Tablet comes bundled with the 129 charging speaker dock for free.

Introducing Google Pixel Fold

In this section, Rick Osterloh introduces the Google Pixel Fold, a foldable phone that combines Tensor G2, Android innovation, and AI for an incredible phone that unfolds into an incredible compact tablet.

Features of Google Pixel Fold

• The device is ultra-premium and costs $1799.

• It is the thinnest foldable on the market.

• The device has three times the screen space of a normal phone.

• It has a custom 180-degree fluid friction hinge.

• The device is built to last with extensive testing done on its hinge to be the most durable of any foldable. Corning Gorilla Glass Victus protects it from exterior scratches while IPX8 water-resistant design safeguards against weather.

Camera System

• Pixel Fold gives you entirely new ways to take stunning photos and videos with Pixel camera.

• You can put the camera in Tabletop mode to capture stars and use the best zoom on a foldable.

• The unique combination of form factor, triple rear camera hardware, and personal AI with Tensor G2 make it the best foldable camera system.

Dual Screen Interpreter Mode

• Your Pixel Fold can use both displays to provide live translation to you and the person you're talking to.

Android Improvements for Large-Screen Devices like Tablets and Foldables

Dave Burke talks about Android improvements for large-screen devices like tablets and foldables.

Continuity Feature

• Continuity is a feature that allows apps to seamlessly adapt from the small screen to the larger screen.

Split Screen Mode

• You can use split-screen mode by swiping up and dragging Google Messages to the side.

• The new Jetpack drag and drop library supports this feature in a wide variety of apps, from Workspace to WhatsApp.

Improvements Throughout the OS

• There are many improvements throughout the OS to take advantage of the larger screen, such as new split keyboards for faster typing and a new two-panel shade showing both notifications and quick settings at the same time.

• Pixel Fold is great for productivity on-the-go with multitasking windows paired.

Google I/O 2021 Keynote

This is the keynote for Google I/O 2021, where they showcase their latest products and technologies.

Pixel Fold: A New Foldable Phone/Tablet

• The Pixel Fold is a new foldable phone/tablet that allows users to adjust the split to suit the content they are viewing.

• The "Help Me Write" feature in Gmail can be used to quickly compose messages.

• The Pixel Fold has a foldable screen that is a game changer and integrates well with Android.

• YouTube enters Tabletop mode when the device is folded at right angles, allowing for an awesome single-handed lean-back experience.

• Support for switching displays from within an app will be added soon.

• The Pixel Fold's camera has a new button on the bottom right that allows users to move the viewfinder to the outside screen, enabling high-quality ultra-wide selfies.

Pre-ordering Pixel Fold and Getting a Free Pixel Watch

• Pre-orders for the Pixel Fold start today and it will ship next month.

• When pre-ordering a Pixel Fold, customers will also receive a free Pixel Watch.

AI Technology Shift

• AI technology shift is as big as they come.

• Google's deep understanding of information combined with generative AI capabilities can transform Search and all of their products yet again.

• No one company can do this alone; developer community will be key to unlocking enormous opportunities ahead.

Conclusion

• The keynote ends with a trailer for the new I/O Flip game.

Project Starline

In this section, we learn about Project Starline's mission to bring people together and make them feel present with each other even if they are miles apart. The new prototype uses a breakthrough AI technique that learns to create a person's 3D likeness and image using only a few standard cameras.

Breakthrough AI Technique

• A breakthrough AI technique has been developed that learns to create a person's 3D likeness and image using only a few standard cameras.

User Experience

• Users report feeling like they are in front of the person they are talking to when using Starline.

• Users express amazement at seeing their colleagues in 3D for the first time.

Diversity in Tech

This section discusses diversity in tech and how it affects product design.

Lack of Representation

• Black women discuss how few technologies consider identities like theirs.

• Developers discuss the importance of providing representation for underserved communities.

Education Through Games

This section highlights the use of games as an educational tool, particularly for closing the math gap in underserved communities.

Educational Games

• 7 Generation Games creates games and tools to close the math gap in underserved communities.

• Game developers combine history and entertainment to touch people's hearts.

Speech Therapy App

In this section, we learn about an app designed to help people with speech impediments access speech therapy globally.

Global Access to Speech Therapy

• Developers discuss the global problem of access to speech therapy.

• The app is designed to help people with speech impediments control their stuttering.

I/O Flip Card Game

This section introduces I/O Flip, a classic card game with a modern AI twist, powered by Google and built for I/O.

Introduction to I/O Flip

• I/O Flip is a classic card game with a modern AI twist, powered by Google and built for I/O.

• Players build their own team, customizing their cards with classes and special powers along the way.

English Choosing a Class and Special Power

Craig chooses pirate as his class and Chloe learns how to tell if someone is a pirate. They then choose their special powers from a pack of 12 cards.

Choosing a Pirate Class and Special Power

• Craig chooses pirate as his class.

• Chloe learns how to tell if someone is a pirate.

• They each get 12 cards in a pack to choose their special powers.

English Creating I/O Flip with Flutter

The team discusses the creation of I/O Flip, including the use of AI tools for image creation, Flutter for animations, Firestore for gameplay communication, and Cloud Run for deployment.

Creating I/O Flip with Flutter

• Maker suite helped prototype prompts while PaLM API generated unique descriptions.

• Animations are powered by code instead of video assets, making them flexible.

• All images were created using Google AI tools trained by artists.

• Firestore was used for gameplay communication like matchmaking and results.

• Cloud Run was used to deploy and scale the all-Dart back end.

English Playing I/O Flip

Craig and Chloe play I/O Flip, discussing its fast-paced nature, customization options, and ability to play quick games or extend winning streaks.

Playing I/O Flip

• The game is fast-paced with customizable cards and characters.

• Quick games can be played when short on time or players can extend winning streaks.

• To play I/O Flip, go to flip.withgoogle.com on a laptop or desktop.