Introduction to Google’s New AI Tools
Google has unveiled a series of new artificial intelligence tools, including a personal AI assistant named Gemini Spark, which is designed to help users by performing tasks proactively. This announcement was made during the company’s annual developers conference, Google I/O, held on Tuesday.
Gemini Spark, part of the emerging trend of ‘agentic’ AI, aims to enhance user productivity by automating routine tasks. Google CEO Sundar Pichai highlighted the potential of such AI agents while addressing attendees at the event in Mountain View, California.
AI Development and Financial Growth
Google and its parent company, Alphabet Inc., have invested billions in artificial intelligence research and development. The company’s chief financial officer indicated that capital expenditures could reach up to $190 billion this year, reflecting a strong commitment to AI advancements. These investments have resulted in notable growth in quarterly earnings, contributing to an 11% increase in stock value since the last earnings report.
Pichai also reported that the Gemini app has seen user engagement surge, with monthly active users increasing from 400 million to over 900 million within a year.
Introduction of Gemini 3.5
The latest iteration of Google’s AI models, Gemini 3.5, is now rolling out to billions of users. The Gemini 3.5 Flash model prioritizes speed and is touted as the most advanced agentic and coding model developed by Google to date, being four times faster than some competing technologies.
This model will become the default for the Gemini app and the ‘AI mode’ on Google Search. Google is also preparing to launch the Gemini Pro version next month, which is currently in internal testing.
Gemini 3.5 has been enhanced with advanced safety features to minimize the risk of generating harmful content or mistakenly rejecting safe queries.
New Features of Gemini Omni
Google has introduced a new model, Gemini Omni, which allows users to generate high-quality videos through various input types, including text, images, videos, and audio. The resulting videos can be easily edited via conversation with the AI model. Future updates will enable the creation of images and audio, though specific timelines for these features remain unspecified.
According to Google, videos produced by Omni will appear more realistic, owing to its understanding of physical forces such as gravity and fluid dynamics.
The Gemini Omni Flash model is launching for Google AI Plus, Pro, and Ultra subscribers through the Gemini app, and it will be available at no cost on YouTube Shorts and the YouTube Create App starting this week.
All videos created using Omni will include a digital watermark called SynthID, and Google plans to implement content credentials verification within the Gemini app. This feature will help users determine whether content was generated by AI or captured with a camera and subsequently edited using AI tools, with a rollout planned for Chrome in the coming months.
Gemini Spark: A 24/7 AI Assistant
Gemini Spark, powered by the 3.5 model, will assist users with mundane tasks such as organizing meeting notes, emails, and chats, ultimately generating documents summarizing key points and action items. Unlike traditional agents, Spark operates in the cloud, allowing it to continue functioning even when devices are inactive.
To address concerns regarding the technology’s capabilities, Gemini Spark is designed to request user permission before executing critical tasks such as sending emails or making purchases. The beta version will be available to select testers starting Tuesday, with plans to expand access to U.S.-based subscribers of the Google AI Ultra tier.
By this summer, Gemini Spark will be integrated directly within the Chrome browser.
Enhanced AI Capabilities in Search and Shopping
Last year’s Google I/O conference saw the introduction of ‘AI mode’ on Google Search, which provides users with conversational responses before displaying relevant links. Since its launch, the number of AI mode queries has more than doubled each quarter, reaching over 1 billion monthly users, according to Google’s head of search, Liz Reid.
The new default model for search will now be Gemini 3.5 Flash, and Google is launching an intelligent search box that adapts to longer queries and offers AI-powered suggestions to assist users in formulating their questions.
Additionally, Google introduced the Universal Cart, described as a ‘truly intelligent shopping cart’ that allows users to add items while browsing, chatting with Gemini, or watching YouTube. This cart will leverage Gemini models to find deals, track price history, and notify users when items are restocked.
The Universal Cart will be available this summer on search and the Gemini app, with future support for YouTube and Gmail.
