Gaming

Gemini 2.0 adds advanced reasoning capabilities to the era of AI Agents

Geeknetic Gemini 2.0 adds advanced reasoning capabilities to the era of AI Agents 1

The artificial intelligence continues to evolve, the languages ​​that are behind all those new features that we can see thanks to the AI They are also moving forward to offer more and more possibilities. Google has presented Gemini 2.0the evolution of your AI that now comes to adapt to the era of AI agents. Gemini version 2.0 is available for developers and will soon be integrated with other Google products.

Google Gemini 2.0 arrives with news in multimodalityis now able to natively generate images or audioyou can also use tools natively. These advances come closer to Google’s vision to create a unique universal assistant. While we will have to wait to test Gemini 2.0, a version experimental Gemini 2.0 Flash will be available to all Gemini users. They will also test the new function Deep Researchwhich grants advanced reasoning abilityexplore complex topics and create reports even with large topics.

Search with AI is also advancing, with Gemini 2.0 it will allow add to the AI search advanced reasoning capabilities that will even allow multi-step questionsuseful for solving equations, for example. This new feature has already begun testing on a limited basis, and hopes to reach more countries next time. 2025.

The progress to reach Gemini 2.0 has been thanks to the experience during this time developing Gemini 1.0, together with dedicated hardware like its Trillium TPUs sixth generation. Under these TPUs all training and inference for Gemini 2.0 has been processed.

You can now try Gemini 2.0 Flash from the Gemini application.

With the launch of this new version of AI, the Google Project Astra has also improved. Now you can offer best dialogues with the possibility of conversing in several languages ​​and even mix thembetter understanding the words they contain accents and the less used. You can also make use of tools like Google searchmake use of Google Lens or Maps. has been added more memory capacityable to remember 10 minutes of the conversation and more past conversations. All this with a reduced latency similar to that of a conversation between humans.

He Mariner Project has been built thanks to Gemini 2.0, which allows early research and analyzes the future of interactions between people and agents. Through the browser you are able to understand the elements on the screen such as text, code or images and forms and can complete the tasks. In the early phase, Project Mariner is not always accurate, but they are working to increase its effectiveness and speed.

Project Mariner has also been built for offer securitywhere he always asks for the user authorization before completing tasks such as purchasing items. At the moment it is testing on a small number of users through an experimental extension in Google Chrome.

With this novelty, they have also been added advances in the world of video games. Thanks to Gemini 2.0 agents will be able to help you guide you through virtual worlds and even reason about the game thanks to the recognition of elements on the screen, which will be able to offer suggestions to continue. In collaboration with Supercell Google is testing the operation of these agentswho look for the answer to rules and challenges in games like Clash of Clans or Hay Day.

These agents also have a role in the real world thanks to the roboticswho wear the reasoning ability of robots.

Some advances that they do not leave safety and responsibility behindwhere work is being done at the same time so that all these functions reach users in a way safe and responsible.

End of Article. Tell us something in the Comments!

Source link

About the author

Redaction TLN

Add Comment

Click here to post a comment