Simultaneous translation into Meet, creation of audio video and smart glasses

Google has hit the table. The Mountain View company has celebrated its annual developer event, Google I/O, where it has presented a flood of novelties that will mark the future of the company, all under the same common denominator: artificial intelligence. If a week ago he demonstrated how AI would be integrated into its Android devices, now Gemini extend to create a universal assistant of AI.

Google’s ambition is clear: if until now it has been synonymous with responding to what the user needed to find, now he wants to go one step further being able to deal with more complex requestspersonalized and integrated into the day -to -day user. He does it with functionalities that understand the context of the user, which surrounds him, follows the conversation, fits the needs beyond the requests and for that he is able to use his full potential.

The starting point is Gemini’s improvement, with progress in versions 2.5 Flash and Pro, as well as introduce deep thinking with greater reasoning in mathematical and code tasks. However, among the great novelties that will allow the integration of AI in the daily life of the user is a change in The search engine that incorporates a new AI mode —The Overviews far beyond Ai. Improvements in the customization of answers, automatic audio translation on Google Meet, monitoring of products and purchases on the web, improves in the creation of images and videos – now with sound – as well as a new look at Android XR, which takes Gemini to glasses and helmets of augmented reality.

“There is more intelligence available, for all, everywhere. And the world is responding, adopting the AI ​​fastest than ever … All this progress means that we are in a new phase of the transition of the AI ​​platform. Decades of research are realized for people, companies and communities around the world “explained Losco Pichai, CEO of Google and Alphabet on stage.

Gemini as universal assistant

Google’s ambition with Gemini is to accompany the user in his day to day. Everywhere, at all times. That is that partner to trust everything we need. “We are building a more personal, proactive and powerful, That enriches our lives, accelerates scientific progress and marks the beginning of a new golden era of discovery and astonishment, “explained Demis Hassabis, CEO of Google Deepmind.

Hassabis comments that they have been working to improve and expand their best multimodal base model, Gemini 2.5 Pro, for make it a “world model”an AI that is able to make plans and imagine new experiences through the understanding and simulation of aspects of the world, “as the brain does.”

This Gemini’s ability to use the knowledge of the world and reasoning to represent and simulate natural environments, It can be used from generating photographs and videos to be incorporated into robots, but will be especially useful in Project Astra, the project that will mark a new era in smart glasses. Here video understanding, screen shared and memory is key.

The arrival of Gemini Live has served Google to improve and explore innovations that allow its AI to relate to the environment, so they have added voices and a more natural and fluid conversation. The idea is that Gemini’s application becomes a universal assistant of AI who will perform daily tasks for us, will be in charge of routine administrative tasks, will present recommendations, making us more productive and enriching the experience.

Here Project Mariner becomes special importance, a research prototype that explores the future of human-agent interaction, starting with browsers. It is a pilot test in the US with a system of virtual agents that can complete up to ten different tasks at the same time. These agents can help To look for information, make reservations, buy or investigate at the same time.

A new search engine: it comes to Mode

But if there is an iconic Google service that will change with the arrival of AI ESA is Search, its search engine. While Chatgpt has proven not to be able to double Google’s arm, the Great g has presented an ambitious plan to completely transform the experience of Find information in Search.

Google explains that since the launch of AI Overviews, there has been a change in users’ requests “Ask more questions, including more complex, long and multimodal questions”which has led to 10% of the use of Google. Now, they go one step further and have launched the IA mode, which will at first will reach the US.

The AI ​​mode is a new search system that Gemini uses with a more advanced, multimodal and powerful reasoning to deepen monitoring questions and links to the web. Use your query branching technique, breaking your question in sub -themes and issuing a multitude of consultations simultaneously on the name of the user. “This allows The search is more on the web than a traditional search on Googlehelping you to discover even more than the web has to offer and to find incredible and hyperrelevant content that matches your question, “explains Liz Reid, head of Google Search.

Among the novelties of Search will be the ability to perform deep searches for investigations, consulting hundreds of searches at the same time and analyzing the results in real time to combine it in an answer – with access to the sources. They can also do Real -time searches with Search Live, A system similar to Gemini Live but that will make the user experience more fluid since you can use the camera of your smartphone.

It will also be centered on purchases. We can ask the search engine “Buy two tickets for this afternoon’s concert”, he will look for what are the best options and show them to the user. It will do it in a deeper way thanks to the alliance with companies such as Ticketmaster, Stubhub, Resy or Vagaro. In addition, Search incorporates a product monitoring of the products, so that if we are interested in a bag, it will be able to warn us when it is at its minimum price and Buy it directly with a touch thanks to Google Pay.

One of the novelties in purchases is the use of artificial intelligence for Try us Clothing virtually, which serves to get an idea of ​​how that pants that interest us from an online store can be left. It will be enough to take a picture for the user to know How that garment can be in your body.

The glasses revolution

In Google they have been working on the concept of smart glasses for more than a decade and it is now with Android XR when they have taken a big step forward. After the announcement of his alliance last year with Samsung in his helmet that will rival Vision Pro, Google goes one step further and has announced an agreement with Gentle Monster and Warby Parker, for Create glasses with Android XR. And the company is clear: “The glasses are only really useful if you want to wear them all day.”

These glasses are equipped with camera, microphones and speakers and work together with your phone, allowing you to access your applications without having to search in your pocket. In addition, an optional screen in the lens provides useful information privately just when you need it. “To the combine these glasses with Gemini, Come and listen to what you do, so that they understand your context, remember what is important for you and provide you with information just when you need it. “It is a direct attack on the finish line.

Although there is still no date of when they will arrive this year, today’s advance has shown how the Android XR glasses will work in real situations, including sending messages to friends, arrange appointments, ask for step -by -step indications, take photos and more. We even made a live translation demonstration between two people, showing the potential of these glasses to break the linguistic barriers, or offering subtitles for the real world. At the moment we will have to wait a few months to see them finished.

Professional video with text

Among Gemini’s improvements is the progress in the generation of images and Videos with Flow, A generative video aspiration model made to see 3, thus enhancing its realism and its fluid dynamics, which have made it a reference of the sector. Now, it also incorporates sound, making the experience total.

Flow’s grace is that the creator can use his own resources to create characters or use flow to create your own images with the text functions in image image. Once a subject or scene is created, They can be integrated into different clips and scenes consistently. Or you can use the image of a scene to start a new shot.

Among the functions that feed this new narrative is a range of functions for professionals or beginners; Chamber control, being able to ask the movement to the movement, The angles and the perspectives of the camera; create scenes and expand existing shots in real time; manage indications much easier; as well as Flow TV, where the user can be inspired by a selection in constant growth of clips, channels and content generated with VI.

To use flow with see 3 will be necessary Have Google Ai Ultra, which will cost $ 249.99 per month and also includes Gemini App with 2.5 Pro Deep Think and I see 3, whisk, noteboollm with the highest limits, Gemini in Gmail, docs and videos, Gemini in Chrome, Project Mariner, YouTube Premium and 30 TB of storage.

Google Beam: Royal and fluid video calls

After presenting it a couple of years ago, Google does not forget Project Starline, its solution to perform video calls approaching telepressure. Called as Google Beamthis platform will use AI to enable “a new generation of devices that help people establish significant connections, regardless of where they are.”

The grace of this system is the use of volumetric video that transforms 2D video transmissions into realistic 3D experiences, which allows to connect more naturally and intuitively. “The combination of our AI video model and our light field screen Create one sensation of dimensionality and depth. This allows to establish visual contact, interpret subtle signals and generate understanding and trust as if they were face to face. “For the development of these devices they are working in collaboration with HP to launch the first Google Beam devices to the market with selected customers at the end of this year.

This solution is not intended for the home, but for offices and work environments. Even so, Google has incorporated one of the functions that will also arrive in Beam in Google Meet; The simultaneous translation into video calls. At first it will be Spanish-English and will allow to maintain authentic translated conversations, almost in real time, keeping the voice, tone and expressions. “It is the first step of a broader vision to see and perfectly understand anyone around the world,” explains the company.