Key Takeaways
- Google’s Gemini AI is getting major upgrades, including new “Live” features for real-time interaction.
- New AI models, Imagen 4 and Veo 3, promise more detailed images and smoother videos with sound.
- AI-powered Search will become more interactive, visual, and personalized for users.
- Google introduced its Android XR platform for smart glasses and headsets.
- The company is exploring “agentic AI” that can complete multi-step tasks for users.
- A new 3D teleconferencing technology called Beam aims to make video calls more lifelike.
Artificial intelligence took center stage at Google’s recent annual developer conference, I/O 2025, which began on May 20.
The company highlighted significant enhancements to its Gemini AI models. The Gemini app will soon let users have real-time conversations about what’s on their screen or seen through their camera, a feature called Gemini Live. This will start rolling out to Android and iOS users this week.
Gemini is also becoming more integrated with other Google apps, allowing users to get directions or set reminders through its conversational interface. Plus, its Deep Research mode is being improved to handle uploaded files and images for more detailed answers.
Google also introduced two new creative AI models. Imagen 4 aims to generate images with much greater detail and better text rendering. Veo 3, a text-to-video model, is designed to create fluid motion graphics and even generate matching soundtracks.
Search is also getting an AI boost. According to Business Standard, the AI Mode in Search will offer more interactive and visually rich results, like AI-generated graphs for complex topics. For shopping, it might offer virtual “try it on” features for clothes. This updated AI Mode is rolling out in the US now, with global plans to follow.
The tech giant also gave a peek at its Android XR platform, designed for mixed reality experiences on smart glasses and headsets. Demonstrations showed smart glasses used for messaging, navigation, and real-time language translation. Google is partnering with eyewear brands like Gentle Monster and Warby Parker for these devices.
A big theme was “agentic AI”—systems that can manage multi-step tasks for you. Google showed off its experimental Project Mariner, which could book tickets or make online purchases entirely through AI. The Gemini app will also get an Agent Mode to retrieve and summarize web data, simplifying complex tasks.
Finally, Google previewed Beam, a new 3D teleconferencing technology. Using a special camera array and display, Beam creates a realistic 3D representation of the person you’re calling, aiming to make virtual meetings feel more like in-person ones. Google is working with partners like Zoom and HP to bring Beam to businesses.
