Search is changing thanks to artificial intelligence

Google wants artificial intelligence (AI) to be beneficial to everyone, and with that goal in mind, it made a series of announcements as part of this year’s Google I/O, which was held in California. The ads focus on Gemini, Google’s largest and most capable AI model.

A year ago, Google announced from the I/O stage its plans to twina family of multimedia visualization artificial intelligence (AI) models capable of reasoning using text, images, video, icons, and more.

At this year’s Google I/O, Google showed how it’s fully operational in the Age of Gemini, showcasing innovative AI solutions in products, research, and infrastructure, and how this brings us closer to the ultimate goal of making AI useful for everyone.

All Google products with over 2 billion users as announced are built on Gemini technology.

The event focused on how to help citizens create new experiences and make Google products more useful.

Google announced, among other things, directly addressing the general public:

* Expanding the overview of artificial intelligence for research. With the new custom Gemini model — capable of multi-step, design, and multimedia thinking — combined with best-in-class search, you’ll soon be able to ask complex multi-step questions, customize search results, and even ask questions with video

* Ask the pictures. More than six billion photos are uploaded to Google Photos every day. With Gemini’s multimedia capabilities, we’re redefining how you search for your photos and videos. Do you want to find a specific memory or remember information hidden in your collection? Just ask ask pictures.

* New ways to interact with Gemini in the workspace. Gemini features will be expanded to more users and integrated into the Gmail sidebar, Docs, Drive, Slides, and Sheets. Additionally, Gemini features will also be added to the Gmail mobile app.

* Gemini for Android. AI technology is integrated directly into the Android operating system. Students can now get help with their exercises by circling problems using Circle to Find. The Gemini overlay will provide dynamic suggestions about what appears on your screen. For example, you can ask to have the PDF summarized for you or use the “Ask this video” option. Additionally, TalkBack with Gemini support will be able to provide more detailed descriptions of images.

Google also announced that it will launch “ Gemini 1.5 Pro For Gemini Advanced subscribers in more than 35 languages, along with a set of 1 million text character sequences that LLMs convert to digital representations.

Gemini 1.5 Pro is the largest of the chatbots widely available to consumers around the world. This means it can understand more information than ever before, such as a 1,500-page PDF file, and will soon be able to understand 30,000 lines of code as well as a one-hour video.

Gemini Advanced subscribers will soon have access to Live, a new mobile chat experience. “With Live, you’ll be able to talk to Gemini and choose from a wide range of different, natural-sounding voices. You’ll also be able to speak at your own pace, with the ability to interrupt with questions, making conversations more understandable.

It was also announced that Google is working closely with the creative community to explore how generative AI can better support the creative process and to ensure AI tools are as useful as possible at every stage:

* foot Show, The model most capable of creating high-definition video, and Imagen 3, the model with the highest quality text-to-image conversion. Additionally, some new recordings – featuring participating artists from around the world – were created using the Music AI Sandbox.

Of course, these advances in AI can only be achieved through truly innovative technological infrastructure. Training modern models requires a lot of computing power.

* foot trillium, 6th generation TPU, which delivers a 4.7x improvement in computing performance per chip over the previous generation, TPU v5e. It will be available to cloud customers later in the year.

A key element of bold innovation is responsible innovation. Thus, we are developing a cutting-edge technology called “AI-assisted red teaming” based on innovative gaming features developed by Google DeepMind, such as AlphaGo, by extending our watermarking technical innovations, such as SynthID, to two new formats – text and video – so that it is easier Discover AI-generated content.

“With the power of Gemini, we aim to make AI useful for everyone. Google’s mission is to organize the world’s information from every source, make it accessible by any means, and combine the world’s information with yours in a way that’s truly meaningful to you. Gemini will help us get there.” , Google notes in its announcement.