Google is advancing its AI technology with the introduction of a new generation of its AI system, Gemini. This system is designed to function as an autonomous assistant capable of performing specific tasks. Google announced this development during the anniversary of the first Gemini launch in Mountain View. One potential application of this AI is to independently find and add items needed for a hobby project to an online shopping cart. However, the actual purchase would still require user confirmation.
Google CEO Sundar Pichai explained that the first generation, Gemini 1.0, focused on organizing and understanding information. With Gemini 2.0, the goal is to make the AI more useful. This new AI assistant is part of Project Mariner, which was implemented with the new generation of AI. According to Google manager Tulsee Doshi, Gemini 2.0 is designed for the era of agents, capable of using intelligent tools and accessing Google products like search. It can even execute program code, enabling these agents to think, remember, plan, and take actions on behalf of users.
These agents can interact with a browser just like a human user, performing actions such as clicking, typing, and scrolling. However, the system also identifies tasks that should not be performed without user consent. For instance, Mariner will not complete a purchase without asking the user first. Initially, Mariner will be tested with trusted users before being made available to the broader public.
In parallel, Google is working on Project Astra, an initiative based on Gemini 2.0 to explore the environment. This project was first introduced at the Google I/O developer conference. The aim is to provide users with useful information not just on smartphones but also through smart glasses. These glasses are similar to the Ray-Ban glasses from Meta, capable of recognizing structures or artworks and assisting with tasks like cooking.
For software developers looking to offer solutions based on Google AI, the Gemini Flash system variant is particularly relevant. It can operate not only on large data center computers but also locally on personal computers or specific smartphone models. On December 11, Google presented Gemini Flash 2.0, which reportedly offers improved performance with similarly fast response times.
Google’s new AI search capabilities have faced some challenges. Images related to this can be viewed in a gallery.
Stay updated on software and development news by subscribing to our newsletter. Please ensure you provide a valid email address to receive updates. If you encounter any issues with the subscription form, try again. Make sure to confirm your subscription via the link in the confirmation email. For more information about the newsletter, click to learn more.