Google I/O: The most customized artificial intelligence (useful) to access Gemini and applications – computers
Introduction from Google I/O was created by 3D images in a film created by artificial intelligence created by the VEO model. The creative ability to mix rainbows, with animals, balloons and gums, which shows the artistic diversity of the artificial intelligence model is impressive. This is moved to the right at Mountain View, Google headquarters, with the Sundar Pichai offer, CEO of the Giant Technology Company.
It indicates that since the last introduction/direction, Google has launched more than 20 copies of artificial intelligence models, which are at the top of the different categories. Gemini 2.5 Pro is the distinctive model is the fastest. Gemini came to the end of The Blue Pokémon, and he got a “affectionate title” from API: artificial Pokemon intelligence.
See the exhibition photos from Google I/O 2025:
Sundar Pichai says that last year he filed a lawsuit against 9.7 billion symbols and this year is already in 480 billion symbols50 times an increase. The research also increased, with more than 1.5 billion reconnaissance with the help of AI Tolly.
One of the technologies provided is Google Beam, which is a system that is able to use different two -dimensional rooms to create a 3D image From the user, in real time. The system will be presented within a few months.
Gmail agents help respond to the user
Google’s translation has also evolved, while maintaining a liquid conversation between different languages, available on Google Meet, Currently between the English and Spanish language for the service subscribers. Gemini Live is now available on iOS and Android Smart Home.
For Project Mariner, Google suggests a model with advanced tools that can take the user’s work. The multi -task capacity now allows you to get 10 simultaneous applications. The system learns to use the user and will soon reach the Gemini Application interface. the The AI Agent Communications System, 2agents agents, already have dozens of partners to prepare their application programming facades. AI agents will start access to Chrome, Search and Gemini.
S. Smart Gmail Steer Customization System, ensures that suggestions are within the same user tone. This is because the system analyzes the way it usually writes in previous emails, tone, elegance and adaptation to the user. The system reaches Gmail during the summer for paid subscribers.
Gemini 2.5 is the most advanced Google
“Gueini 2.5 Flash arrives in early June, with the latest developments in the AI of Google.. The system has the ability to deduct different tones in 24 languages, alternating between languages, and maintaining the same sound. The Gemini applications interface already allows you to test this system.
The spread of Gemini is able to create models more than five times from Gemini Flash 2.0 Lite. This system will reach Gemini 2.5 Flash, able to generate a faster code for those who need to help programming. the Google’s deep thinking mode will help give more reflection answers and says it has already started getting good results in standard tools.
Gamini 2.5 thinking allows you to create videos or even vibrant interactive elements from one image, for example. the Gemini follows robots to teach robots through artificial intelligence.
Project ASTRA helps users access evidence by helping the user to do specialized tasks such as getting a bike. Whether searching in the reform guide, YouTube search solutions, or contacting a store to buy spare parts, the system is to be a productive and useful assistant for everyone. Project ASTRA arrives at Gemini Live and Search Live and Live Api for developers. Project ASTRA has a new project to help people with poor eyesight find an object, and works literally as their notables, describing what you see.
Google lens and search with artificial intelligence mode
Google Lens had a 65 % increase in use and with the help of artificial intelligence, people could find what they were looking for. The new situation allows you to put more sophisticated questions with likely answers. Mode AI today reaches all users in the United States. Google says users change the way they interact with the search. They no longer ask direct questions, but with more detailed descriptions awaiting more full answers as well. Google Search Mode AI, a new tab with the main AI models in Google and its innovations. An artificial intelligence overview has already begun to be a reflection of this new life of research. But technology is more in the future.
An example of this is the ability to help define sporting events, restaurants, etc., with detailed information, prices and ability to buy directly using Google Pay without leaving Google Search. to The capabilities of the ASTRA project will be included in the search called Search Live. The system collects the available information on the web, and expresses what it sees in the actual time.
Gemini on Chrome helps on the Internet to browse the Internet, and is always ready to answer the questions you ask. The job this week arrives for users paid in the United States. The Imagen 4 model comes to Gemini, the most powerful Google to create images. The model can generate more realistic pictures, but it can also be rejected with the creators of users. VEO3 is equal to you allows you to create videos, such as those created to present loudspeakers on the stage and the introduction to the events. The system can add sounds and dialogues now.
Writing Note: An article updated with more information. Last update 20h04.
