GCP, Google Workspace
Google I/O 2024 was a whirlwind of innovation, with a heavy emphasis on AI and the powerful Gemini models. The two-hour keynote, led by CEO Sundar Pichai, showcased a variety of groundbreaking technologies and features that promise to transform our digital lives.
Ask Photos
Google has introduced “Ask Photos,” a new feature in Google Photos powered by Gemini. This feature allows users to search their photo libraries using natural language queries.
Photos launched almost nine years and today, almost 6 billion photos and videos are uploaded every day. With this amount of data, browsing your valuable memories might not always be straightforward. Thanks to artificial intelligence, Gemini simplifies the process, allowing users to easily search their photos and metadata. Users will be able to find information in their galleries with natural language questions, such as “Show me the best photo from each national park I’ve visited” and others.
Gemini AI can detect scams in calls
Google unveiled a new feature that uses its Gemini Nano AI to detect scams during phone calls. Unlike current systems, which only alert users to potential spam when the call comes in, Gemini Nano will listen for scam-like conversation patterns in real time.
For example, if someone pretends to be a bank representative or requests passwords and gift cards, the system will recognize these tactics and notify the user about the potential scam. This feature, built into a future version of Android, will run entirely on-device for enhanced privacy.
New Gemini models
Google introduced the enhanced Gemini 1.5 Pro, capable of processing up to 2 million tokens, doubling its previous capacity and surpassing other commercially available models like Anthropic’s Claude 3. This advancement positions Gemini 1.5 Pro as a leading AI model in handling large-scale data inputs.
Gemini 1.5 Flash
Google also introduced Gemini 1.5 Flash, a streamlined version of its Gemini 1.5 Pro model. Designed for high-frequency, less demanding tasks, Flash emphasizes speed and efficiency while supporting up to 2 million tokens. It retains the multimodal capabilities of the Pro version, analyzing audio, video, and images but generating only text.
Gemini for Google Workspace
Google announced that Gmail is getting an AI-powered boost with Gemini. This upgrade allows users to search, summarize, and draft emails using AI. Gemini can also handle complex tasks, such as processing e-commerce returns by finding receipts and filling out online forms.
In a demo, Google showed how Gemini could summarize emails from a school for busy parents or compare quotes from contractors for home renovations. Gemini can also help organize receipts, extract information into spreadsheets, and automate frequent tasks, making it ideal for business travelers.
A button to activate Google’s Gemini AI will be available in the side panel of various Google apps, including Gmail, Google Drive, Docs, Sheets, and Slides. Google is also adding language support for more Gemini for Workspace features.
Google Veo
Google introduced Veo, an AI model designed to create 1080p video clips up to a minute long from text prompts. Veo can generate various visual and cinematic styles, including landscapes and time lapses, and edit and adjust existing footage. It builds on Google’s previous video generation efforts using the Imagen 2 family of models but significantly improves by offering higher resolution and longer video duration. Veo is positioned to compete with leading video generation models like OpenAI’s Sora.
Project Astra
Project Astra is an advanced visual chatbot that works via voice commands through phone cameras and potentially through smart glasses. Astra seems able to handle all sorts of queries, showing its ability to understand and interact with the environment.
These announcements underscore Google’s commitment to advancing AI technology and providing innovative solutions for businesses and consumers alike. Watch the full keynote here.
