Text by: Google
- Expansion of AI summaries in Search. With a new customized Gemini model — capable of reasoning in multiple steps, planning, and multimodality — combined with top-notch search functionality, you will soon be able to ask complex, multi-step questions, customize search results, and even ask questions with videos [Blogpost]
- Introduction of Ask Photos. More than six billion photos are uploaded to Google Photos every day. With the multimodal capabilities of Gemini, we are redefining how you can search your photos and videos. Want to find a specific memory or retrieve information hidden in your gallery? Just ask Photos. [Blogpost]
- New ways to work with Gemini in Workspace. Gemini's capabilities will be extended to more users and integrated into the sidebar of Gmail, Docs, Drive, Slides, and Sheets. Gemini features will also be added to the mobile app of Gmail. [Blogpost]
- Gemini for Android. We are building AI directly into the Android operating system. Students can get help with their homework by circling problems with Circle to Search. And Gemini's overlay will also provide dynamic suggestions related to what is on your screen — summarizing a PDF or "ask this video" — while TalkBack with Gemini will be able to provide even more detailed descriptions of images. [Blogpost]
We are also bringing Gemini 1.5 Pro to Gemini Advanced subscribers in more than 35 languages, along with a context window of 1 million tokens — the longest of any widely available consumer chatbot in the world. This means it can understand more information than ever before, such as a 1500-page PDF and soon, 30,000 lines of code and a one-hour-long video.
- Gemini Advanced subscribers will also soon have access to Live, a new mobile conversation experience. With Live, you can talk to Gemini and choose from various naturally sounding voices. You can speak at your own pace and even interrupt with questions, making conversations more intuitive. [Blogpost]
And we are also looking forward to the next models: we have shared more details about Gemini 1.5 Flash, a cost-effective, faster model built based on user feedback, and Project Astra, our vision for the next generation of AI assistants, a responsive agent that can understand and respond to the context of conversations. [Blogpost]
We are also working closely with the creative community to discover how generative AI can best support the creative process and to ensure that our AI tools are as useful as possible at every stage:
- We are introducing Veo, our most capable model for generating high-definition video, and Imagen 3, our highest quality text-to-image model. We are also sharing new demo records — with global artists — created with our Music AI Sandbox. [Blogpost]
Of course, these advances in AI are only possible through truly advanced infrastructure technology. Training state-of-the-art models requires a lot of computing power.
- We have unveiled the 6th generation of our TPUs, called Trillium, which offers a 4.7x improvement in computing performance per chip compared to the previous generation, TPU v5e, and which we will make available to Cloud customers later this year. [Blogpost]
Bold innovation must be supported by responsible innovation. That is why we are developing an advanced technique we call AI-assisted red teaming. This technique leverages Google DeepMind's breakthroughs in gaming, such as with AlphaGo, and extends our technical innovations for watermarking - such as SynthID - to two new modalities: text and video. This makes AI-generated content easier to identify. [Blogpost]
With the power of Gemini, we want to make AI usable for everyone. Google's mission is to organize the world's information through any input, make it accessible through any output, and combine the world's information with the information in your world, in a way that is truly useful to you. Gemini will help us achieve that goal.
