Google outlines ambitious future for AI assistant Gemini ahead of I/O 2025

Unlike previous iterations of Google Assistant, Gemini is designed to act without being prompted.

Google is significantly expanding its assistant capabilities through Gemini, a new AI-driven experience designed to be personal, proactive and powerful. The company’s vision was laid out by Josh Woodward, VP at Google, in a recent post on X (formerly Twitter), marking a clear strategic pivot towards a more intuitive, context-aware digital assistant.

Underpinning Gemini’s capabilities are some of Google’s most advanced AI models, (REUTERS)

By Ayushmann Chawla

Ayushmann Chawla is the Deputy Chief Content Producer at Hindustan Times. A seasoned tech journalist with years of experience working for some of the industry’s leading media organizations, his articles can also be read on Live Mint. His passion extends beyond journalism—he’s a dedicated automobile enthusiast, always ready to explore the latest in car technology and design. A self-confessed gadget lover, Ayushmann finds joy in testing new devices and sharing his findings with his audience. Instagram: ayushmannchawla

Read moreRead less

Personalisation at the core

At the heart of Gemini is a focus on personalisation. Internally referred to as “pcontext”, the assistant will use previous interactions and context drawn from across Google's suite of services—including Gmail, Calendar, Search and Photos—to tailor responses and suggestions. This will be done with user consent and data permissions clearly in place.

Proactive assistant behaviour

Unlike previous iterations of Google Assistant, Gemini is designed to act without being prompted. The system will surface useful information, reminders or recommended actions automatically, aiming to anticipate what users need before they ask. This shift is intended to minimise friction and help users stay productive.

Powerful tools backed by DeepMind

Underpinning Gemini’s capabilities are some of Google’s most advanced AI models, including Gemini 2.5 Pro. With support from Google DeepMind, the assistant will be able to perform complex tasks such as generating content, writing code, and managing projects. Google says Gemini is being developed not only to respond to questions, but also to carry out meaningful work on behalf of users.

Expansion and accessibility

Gemini is already being rolled out to students across the United States, with more regions expected to follow. Thanks to Google's infrastructure, particularly its Tensor Processing Units (TPUs), the service is designed to run at scale and, notably, without cost to users.

Rapid updates and new features

Google also announced a series of related updates: the release of Gemini 2.5 Flash, the launch of Veo 2 for video generation, LaTeX support for document creation, and a new image editing and upload tool. Woodward emphasised Google’s rapid development cycle, urging users to share feedback to help shape future updates.