Note from Google and Alphabet CEO Sundar Pichai:
Information is at the core of human progress. That’s why for more than 26 years, we’ve been focused on our mission of organizing the world’s information, making it accessible and useful. And that’s why we continue to push the frontiers of AI, organizing that information across all inputs, making it accessible from all outputs, and making it truly useful for users.
That was our vision when we introduced Gemini 1.0 last December. The first models built to be natively multimodal, Gemini 1.0 and 1.5 are multimodal and capable to understand and process more information across text, video, images, audio, and code. Long contexts have brought about great progress.
Millions of developers are developing with Gemini today. And it also helps us rethink all our products and create new ones, including all seven products with 2 billion users. NotebookLM is a great example of what multimodality and long context can do for people, and why so many people love it.
Over the past year, we have invested in developing more agent-like models. This means your agent has a deeper understanding of the world around you, can think several steps ahead, and can take action on your behalf under your supervision.
Today, we’re excited to announce the next generation model built for this new age of agents. Introducing Gemini 2.0, our most capable model to date. New advances in multimodality and the use of native tools, such as native images and audio output, will allow us to build new AI agents that move even closer to the vision of a universal assistant.
Today we’re putting 2.0 into the hands of developers and trusted testers. And we’re working quickly to incorporate this into our products, starting with Gemini and Search. Starting today, the Gemini 2.0 Flash experimental model is available to all Gemini users. We’re also releasing a new feature called Deep Research. This feature uses advanced reasoning and long-context capabilities to act as a research assistant, researching complex topics and creating reports on your behalf. Currently available in Gemini Advanced.
No product has been more transformed by AI than search. Our AI Overview is now used by 1 billion people, allows you to ask completely new types of questions, and has quickly become one of the most popular search features of all time. The next step is to bring Gemini 2.0’s advanced reasoning capabilities to AI Overview to tackle more complex topics and multi-step questions, such as advanced formulas, multimodal queries, and coding. We are planning to begin limited testing this week with a broader rollout early next year. And we’ll continue to bring AI Overview to more countries and languages over the next year.
2.0 progress is underpinned by a decade of investment in a differentiated full-stack approach to AI innovation. Built on custom hardware such as Trillium, a 6th generation TPU. TPUs power 100% of Gemini 2.0’s training and inference, and now customers can also build with Trillium, which is generally available.
If Gemini 1.0 was about organizing and understanding information, Gemini 2.0 is about making information more useful. I can’t wait to see what the next era will bring.
-Thunder
