Throughout 2023, Google introduced incredible new AI innovations to its customers and the wider community of developers and users: AI Hypercomputer to train and deliver Productive AI models; Productive AI support in Vertex AI, enterprise AI platform Vertex AI; Duet AI in Google Workspace; and Duet AI for Google Cloud. A host of new capabilities in its AI-optimized infrastructure with notable advances in GPUs, TPUs, machine learning software and compilers, workload management, and other areas; prepared a whole new suite of capabilities with many innovations in Vertex AI and Duet AI agents in Google Workspace and Google Cloud Platform.

Google today introduces a number of key new features in its AI product line to support its most capable and generic model ever, Gemini.

It is designed to be multi-modal from scratch, that is, it can generalize and seamlessly understand different types of information, including text, code, audio, image and video, as people see, hear, read, listen and talk about many different types of information at the same time, work and combine them.

‍

Integrated AI support for Google Cloud

As of today, Gemini is part of a vertically integrated and vertically optimized set of artificial intelligence technologies, all of which are made up of several key parts designed to work together:

‍

• Scalable AI infrastructure: Google Cloud offers a leading AI-optimized infrastructure for companies, used by Google to train and deliver models. Our entire AI infrastructure suite was built with system-level code design to increase efficiency and productivity in AI training, tuning, and delivery processes.
• World-class models: Google continues to offer a series of artificial intelligence models with different skills. In late 2022, the Pathways Language Model (PalM) had been launched, soon after, PalM 2 was released, and now Gemini Pro is presented. Domain-specific models such as MED-Palm and SEC-Palm were also made available.
• Vertex AI - the leading enterprise AI platform for developers: Vertex AI, the AI development platform, was rapidly developed to help developers build agents and integrate productive AI into their applications. Vertex AI helps customers discover, customize, augment, deploy, and manage agents built using the Gemini API, as well as a select list of more than 130 open-source and third-party AI models that meet Google's strict enterprise security and quality standards. Vertex AI takes advantage of Google Cloud's built-in data management and privacy controls, and also provides tools to help developers use models responsibly and safely. Vertex AI also provides Search and Speech, tools that use a low-code approach to develop advanced search, and speech agents that can work across many channels.
• Duet AI - Auxiliary AI tools for Workspace and Google CloudDuet AI acts as an AI-powered assistant that helps users use Google Workspace and Google Cloud. Duet AI in Google Workspace, for example, helps users write, create visuals, analyze spreadsheets, draft and summarize emails and chat messages, and summarize meetings. Duet AI in Google Cloud helps users code, deploy, scale, and monitor applications, as well as identify cybersecurity threats and accelerate resolution.

‍

Strengthening our world-class infrastructure

As the size and complexity of AI models increases, so do the requirements for training, adjustment, and inference. As a result, the demand for high-performance, highly scalable and cost-effective AI infrastructure to train and deliver models is growing exponentially.

This applies not only to customers, but also to Google. TPUs have long been the basis for training and delivering AI-powered products such as YouTube, Gmail, Google Maps, Google Play and Android. In fact, Gemini was trained on TPUs and presented using TPUs.

AI Hypercomputer, a groundbreaking supercomputer architecture that uses an integrated system of performance-optimized hardware, open software, leading machine learning frameworks, and flexible consumption models, has also been announced. AI Hypercomputer has a wide range of accelerator options, including multiple 5th generation TPU classes and NVIDIA GPUs.

‍

Presentation of the latest groundbreaking models

Gemini is also the most flexible model that can work efficiently anywhere, from data centers to mobile devices. While Gemini Ultra is the largest and most capable model for highly complex tasks, Gemini Pro is the best model for scaling across a wide range of tasks, and Gemini Nano is the most efficient model for on-device tasks. Its state-of-the-art capabilities will significantly improve the way developers and enterprise customers develop and scale with artificial intelligence.

An upgraded version of Imagen 2 offers enhanced photorealism, text processing, and logo creation, allowing you to easily create text-overlaid images and produce logos.

In addition, Google has enhanced its efforts on domain-specific models with MED-Palm, and announced the MEDLM family of basic models, fine-tuned for healthcare use cases.

‍

‍

Strengthening the Vertex AI platform with Gemini

Today, we are pleased to announce that Gemini Pro is available in preview on Vertex AI. It currently empowers developers to create new and differentiated agents that can process information across text, code, images and video. Vertex AI helps you deploy and manage agents into production, automatically evaluate the quality and reliability of agent responses, and monitor and manage them.

Vertex AI provides comprehensive support for Gemini with the ability to discover, customize, augment, manage, and deploy agents created against the Gemini API.

‍

Expanding the capabilities of Duet AI

With Duet AI, we are committed to helping our customers increase their productivity, gain competitive advantage, and ultimately improve their profitability. Duet AI for Developers helps users write code faster with AI code completion, code generation, and chat across multiple integrated development environments (IDEs). Facilitates repetitive developer tasks and processes with shortcuts for common tasks such as unit test creation and code annotation, speeds up troubleshooting and troubleshooting, and helps reduce context switching. Duet AI also accelerates skill-based learning by giving users the ability to ask questions using natural language chat.

With Duet AI in Security Operations, Google Cloud's unified security operations platform, security teams can leverage the power of productive AI to enhance their skills and help accelerate threat detection, investigation and response. With Duet AI in Security Operations, users can search large amounts of data in seconds with natural language generated custom queries, reduce time-consuming manual reviews, quickly uncover critical context by leveraging automated summaries of case data and alerts, and improve response time using recommendations for next steps to support incident remediation We present me.

Google owns the entire Duet AI technology suite, from infrastructure and core models to high-end integration and user experience.

Earlier this year, BigQuery began working on expanding Duet AI across its portfolio, including Duet AI in Looker, database products, Apigee, and more.

‍

Supporting next-generation AI solutions

The launch of Gemini, together with our super-scalable AI infrastructure portfolio Vertex AI and Duet AI, delivers a comprehensive and powerful cloud for developers and customers. With these innovations, Google Cloud is powering the next generation of AI-powered agents in every industry, enabling organizations to create, use, and successfully adopt AI to support their digital transformation. If you want to get detailed information about the solution group, you can contact us.

‍