“Information is at the core of human progress. It’s why we’ve focused for more than 26 years on our mission to organise the world’s information and make it accessible and useful. And it’s why we continue to push the frontiers of AI to organise that information across every input and make it accessible via any output, so that it can be truly useful for you,” said Sundar Pichai, CEO of Google and Alphabet.
Gemini 2.0 Flash is now available as an experimental model to developers through the Gemini API in Google AI Studio and Vertex AI. Google aims to quickly integrate it into products like Gemini and Search. Starting December 11, Gemini 2.0 Flash will be accessible to all Gemini users.
Google also unveiled Deep Research, a feature leveraging advanced reasoning and long-context capabilities to act as a research assistant. It explores complex topics and compiles reports on behalf of users. This feature is available within Gemini Advanced.
Also Read: Google and Vodafone Expand Partnership to Bring AI-Powered Services Across Europe and Africa
Enhancements in AI Search
AI overviews now reach over 1 billion users globally. Google plans to incorporate Gemini 2.0’s advanced reasoning capabilities into these overviews to address complex topics, multi-step questions, advanced math equations, multimodal queries, and coding challenges. Testing has begun, with broader rollout expected early next year. By 2025, AI overviews will expand to more countries and languages.
“2.0’s advances are underpinned by decade-long investments in our differentiated full-stack approach to AI innovation. It’s built on custom hardware like Trillium, our sixth-generation TPUs. TPUs powered 100 percent of Gemini 2.0 training and inference,” Pichai noted. “If Gemini 1.0 was about organising and understanding information, Gemini 2.0 is about making it much more useful.”
Also Read: Google Announces AI Collaborations for Healthcare, Sustainability, and Agriculture in India
Gemini 2.0 Flash: A Workhorse Model
“We are releasing the first model in the Gemini 2.0 family of models: an experimental version of Gemini 2.0 Flash. It’s our workhorse model with low latency and enhanced performance at the cutting edge of our technology, at scale,” said Demis Hassabis, CEO of Google DeepMind and Koray Kavukcuoglu, CTO of Google DeepMind on behalf of the Gemini team.
Gemini 2.0 Flash
The first model in the Gemini 2.0 family, Gemini 2.0 Flash, is optimised for low latency and enhanced performance at scale. According to Google, it outperforms Gemini 1.5 Pro on key benchmarks while operating at twice the speed. Notably, it supports multimodal outputs such as natively generated images combined with text and steerable multilingual text-to-speech (TTS) audio.
Google is also releasing a new Multimodal Live API, enabling real-time audio and video streaming inputs with the use of multiple combined tools.
Also Read: Google Features Startups Using AI to Transform Mental Health Support
Agentic Experiences
According to Google, Gemini 2.0 Flash’s native user interface action-capabilities, along with other improvements like multimodal reasoning, long context understanding, complex instruction following and planning, compositional function-calling, native tool use and improved latency, all work in concert to enable a new class of agentic experiences.
Google said it is exploring prototypes built on Gemini 2.0, including:
Project Astra: A personal AI assistant with enhanced memory, multilingual dialogue, and integration with Google tools like Search, Lens, and Maps. Astra now retains up to 10 minutes of in-session memory and can recall past conversations. Google plans to extend these capabilities to Gemini and AR glasses.
Project Mariner: A browser-based agent capable of completing tasks by interpreting web elements and user interactions.
“Project Mariner is an early research prototype built with Gemini 2.0 that explores the future of human-agent interaction, starting with your browser. As a research prototype, it’s able to understand and reason across information in your browser screen, including pixels and web elements like text, code, images and forms, and then uses that information via an experimental Chrome extension to complete tasks for you,” Google explained.
Jules: An AI-powered coding assistant integrated with GitHub workflows to support software development. This effort is part of Google’s long-term goal of building AI agents that are useful across all domains, including coding.
Gaming and Robotics
Gene 2 Launch
At this juncture, Google has also highlighted the launch of its large-scale foundation world model, Genie 2, which the company unveiled on December 4. Genie 2 is a foundation world model capable of generating an endless variety of action-controllable, playable 3D environments for training and evaluating embodied agents. Building on this advancement, Google said it has built agents using Gemini 2.0 that can help users navigate the virtual world of video games.
Also Read: Google AI Innovations: Key Announcements From October and November 2024
Robotic Applications
Beyond virtual applications, Google is experimenting with agents that apply Gemini 2.0’s spatial reasoning capabilities to robotics, enabling new possibilities in the physical world.
“Today’s releases mark a new chapter for our Gemini model. With the release of Gemini 2.0 Flash, and the series of research prototypes exploring agentic possibilities, we have reached an exciting milestone in the Gemini era,” Google said on December 11.
The company plans to integrate Gemini 2.0 across its suite of products, starting with Search and the Gemini app, while continuing to explore its capabilities in collaboration with developers, trusted testers, and experts.
Some people read for free. A few choose to support. If you found TelecomTalk useful, you can help keep us running.
Support Telecomtalk
