Google strikes back: AI takes center stage at I/O 2025

In 2023, Microsoft CEO Satya Nadella had confidently stated that his company’s AI lead would make Google “Dance.” Fast forward to May 20, 2025, and Google has responded in dramatic fashion at its annual I/O conference, showcasing a wave of powerful AI advancements. This year’s event was a clear statement of intent, less about Android updates and more about how Gemini, Google’s flagship AI model, will redefine search, creativity, and device interaction.

Also Read:Look to scroll: Apple’s Vision Pro may soon let your eyes do all the work

At the heart of these announcements is Gemini 2.5 Pro, the latest and most capable version of Google's language model. It introduces a new “Deep Think” mode, which gives the AI enhanced reasoning skills by dedicating more computational resources to analyze complex problems. This allows it to approach tasks like coding and math with a level of depth previously not possible, almost replicating human-like logical thinking.

Google Search is also undergoing a radical transformation. The new AI Mode lets users pose more detailed and nuanced queries. Instead of the usual list of links, the system offers AI-generated summaries and contextual insights, making the search experience more conversational and intuitive.

In a surprising move, Google is revisiting smart glasses with Android XR, a new operating system designed for immersive wearables. Partnering with brands like Gentle Monster and Warby Parker, the goal is to integrate Gemini AI into glasses that can interpret your surroundings, read messages, and provide real-time information, all right in front of your eyes.

In creative AI, tools like Imagen 4 and Veo 3 got major upgrades. Imagen now produces better images and readable text, while Veo generates videos with matching audio. The new Flow tool helps creators extend scenes and maintain consistency across projects, signaling a potential game-changer for filmmakers and developers.

Lastly, Gemini Live, enhanced with Project Astra, brings real-time visual and audio assistance to mobile devices. It can understand what's around you, detect emotions in your voice, and respond accordingly, all while ignoring background noise.