On May 20, 2025, the annual Google I/O 2025 developer conference kicked off in Mountain View, California. The event showcased numerous updates in the fields of artificial intelligence and communication technologies.
This is reported by Business • Media
New Google Beam Technology and Improvements to Google Meet
Among the key innovations at the conference was Google Beam — a video communication platform that provides a realistic 3D presence effect. Utilizing six cameras and artificial intelligence, Google Beam offers playback with millimeter accuracy at 60 frames per second, creating the illusion of conversing in the same room. The first Beam devices are expected to be released for early users in 2025.
Additionally, Google Meet has been enhanced with real-time speech translation. This feature reproduces the speaker’s voice, tone, and facial expressions, currently working from English to Spanish in beta for users of the AI Pro and Ultra plans, with plans to add more languages in the future.
Generative Models Veo 3 and Imagen 4
The conference also introduced the new generative model Veo 3, capable of creating videos that include sound effects, background noise, and dialogues. This model has significantly improved video quality compared to previous versions and will be available starting May 21 in the Gemini app for subscribers of the AI Ultra plan at $249.99 per month.
Furthermore, Imagen 4 — an image generation tool that can create photorealistic and abstract images with high detail, including textures and natural elements, at resolutions up to 2K.
The event also announced updates to Gemini Live, which introduced features of the universal AI assistant Project Astra with camera support and screen sharing, already utilized for interview preparation and marathon training.
Google continues to develop AI-based agents that automate routine tasks. The Mariner prototype allows users to conduct housing searches and other tasks, available to developers through the Gemini API.
“If a friend asks for travel advice, Gemini will automatically find your previous emails and files, such as itineraries in Google Docs, and suggest a response that matches your style and tone,” Google stated.
Additionally, Google announced a new AI Mode that allows for longer and more complex queries in search, as well as an updated Deep Research feature that now supports uploading personal PDF files and images.
In conclusion, Google presented Stitch — a tool for generating web and mobile app interfaces that creates HTML and CSS code based on specific words or images.
The Google I/O 2025 conference also included an announcement of a $150 million investment in eyewear manufacturer Warby Parker, highlighting the company’s commitment to innovation across various sectors.