Google I/O 2024: From Gemini 1.5 Flash to new AI features in Android, check out the top announcements

At its I/O 2024 event on Tuesday, Google unveiled several significant upgrades, including the release of new Gemini and Gemma models, advanced AI features for Android, a new AI voice assistant, and a text-to-video generator.

Gemini 1.5

Google introduced the Gemini 1.5 Flash language model, which is more lightweight, fast, and efficient than the popular Gemini 1.5 Pro model. Capable of multimodal reasoning, it features a larger context window of 1 million tokens and is optimized for high-frequency tasks requiring quick response times.

Gemma

The next generation of the Gemma AI model was revealed, boasting improved performance and efficiency, optimized by TPUs and GPUs, and featuring 27 billion parameters. Google also announced PaLI-3, its first vision language model, as part of the Gemma family, with a release set for June.

Ask Photos

Google Photos is receiving a Gemini AI-powered upgrade called ‘Ask Photos,’ allowing users to recall specific pictures based on a single prompt. Utilizing Gemini’s multimodal abilities, the feature understands the context and subjects in images to quickly locate requested photos.

AI in Android: Google confirmed that new AI capabilities, powered by Gemini, are coming to Android, including the popular Circle to Search feature, which will help students with homework by providing step-by-step instructions for solving physics and math problems. Additionally, Gemini Nano, Google’s lightweight on-device large language model, will gain multimodal capabilities and a new safety feature for real-time scam alert detection during calls. Gemini will also be integrated across different apps, allowing users to interact with the AI assistant directly on their screens.

VEO

Google introduced VEO, a new text-to-video generation model, as a competitor to OpenAI’s Sora. Capable of generating 1080p resolution videos with various cinematic and visual styles, VEO offers users greater control through its understanding of cinematic terms like ‘timelapse’ and ‘aerial shots.’

6) AI Search

Google is enhancing its search experience with a customized Gemini model, providing quick AI-generated summaries for search topics and a new AI-organized search results page. Initially available in the US, this feature will first roll out for dining and recipe searches, followed by categories like movies, music, books, hotels, and shopping.

7) Project Astra

Google introduced Project Astra, an AI agent designed to rival OpenAI’s GPT-4. Demonstrated capabilities include identifying objects in a room, explaining code, determining location by looking outside, finding items like glasses, and generating creative names for pets. Project Astra may lead to a significant Gemini-powered upgrade to Google Lens, with applications showcased on smartphones and smart glasses.

These advancements highlight Google’s commitment to integrating cutting-edge AI technologies across its platforms and services.