Key Highlights of Google Cloud Next '23

2023-08-30

Just three months after OpenAI's ChatGPT made its debut, Google launched Bard. Since then, Google has been riding the wave of generative AI and introducing new features in various fields (including Vertex AI, Model Garden, etc.). Over 70% of AI unicorns operate on Google Cloud infrastructure, such as Anthropic, AI21 Labs, and Character.AI, and Google leverages these partnerships to provide global services.

Following Google I/O Connect, Google Cloud Next will take place in San Francisco from August 29th to 31st. Let's take a look at the key announcements and upgrades from this event.

Vertex AI Gets Bigger and Better

Vertex AI, the cornerstone suite of Google's generative AI products, received new upgrades at Google I/O 2021. Their proprietary LLM PaLM now has a four-fold increase in input token length and supports 38 languages. The new version of their AI coding robot, Codey, has improved code generation quality by 25%. The text-to-image diffusion model, Imagen, also has new updates that enhance image quality and introduce a new feature called "style adjustment".

In addition, Meta's Llama 2, Anthropic's Claude 2, and TII's Falcon have joined Model Garden, expanding the variety of models available. Vertex AI Extensions support model data connections, allowing read-only access to Google and third-party data. Vertex AI Search brings Google search capabilities to business data, and Vertex AI Conversation enables easy deployment of chatbots. Furthermore, Colab Enterprise combines user-friendly Colab notebooks with enterprise-level security. Vertex AI Search is now generally available.

Duet AI for All Users

The good news is that Duet AI in Google Workspace is now publicly available.

In Google Meet, Duet AI can improve your appearance and enhance audio quality. It can also take notes and even attend meetings on your behalf. Duet AI in Google Chat can answer questions, summarize content, and assist in retelling conversations, now with improved UI and connectivity options.

In addition to Workspace, Duet AI in Google Cloud is expanding its capabilities for various user groups, including developers, data practitioners, and cybersecurity professionals. It is integrated into products such as BigQuery, Looker, and Database Migration Service, providing contextual help, automatic suggestions, and collaboration features. The launch of BigQuery Studio provides an integrated interface for data engineering, analysis, and machine learning. Furthermore, Google Cloud is facilitating the integration of Vertex AI models into BigQuery and introducing a downloadable version called AlloyDB Omni that can run on multiple platforms.

Google has also introduced AlloyDB AI, which makes it easy to build enterprise AI applications, and released Cloud Spanner Data Boost for efficient operational data processing. Developer tools like Jump Start Solutions and Application Integration are becoming generally available, and partnerships with GitLab provide support for secure DevOps solutions. The Duet AI toolkit of these extensions includes cloud workstations and application integration support.

Advancements in Google Supercomputing and Cloud Services

Google revealed that their Cloud A3 supercomputer will be fully available starting next month. This supercomputer, based on NVIDIA H100, is designed specifically for demanding generative AI workloads and LLM. A3 leverages Google's infrastructure processing units to provide three times faster training speeds and improved network connectivity compared to the previous generation. Additionally, cross-cloud networking has been introduced as a new platform to ensure secure connections between cloud applications.

Furthermore, with Google Distributed Cloud (GDC), customers can flexibly deploy AI and data workloads using selected Vertex AI services and AlloyDB Omni. Additionally, the advanced GKE Enterprise Edition offers multi-cluster horizontal scaling to meet the demands of rigorous AI and ML workloads.

Prioritizing Sustainability

Google Maps has introduced three environmental APIs - solar energy, air quality, and pollen - aimed at helping businesses reduce emissions and adapt to environmental changes. It caters to various health needs by providing localized pollen count data, heatmaps, comprehensive allergen information, and practical strategies to reduce pollen exposure.

The APIs calculate daily pollen particle levels on a 1x1 square kilometer grid in over 65 countries and can forecast for the next five days, including three plant categories and 15 species. By integrating factors such as land cover, climate data, and pollen production, the model can provide reliable predictions of local pollen levels and associated risks.

Expanding Partnerships

Google Cloud is collaborating with a range of customers and organizations, from startups to enterprises, to enhance their generative AI capabilities. A121 Labs is integrating industry-specific generative AI into BigQuery, Bayer is using Vertex AI Search and Med-PaLM 2 for medical purposes, and Culture Amp is utilizing Vertex AI for employee feedback.

MSCI is helping investors with risk management, Runway is deploying AI for creative processes, and Six Flags is implementing AI for park guidance. Google Cloud has also expanded its partnership with SAP and Docusign in industry solutions and contract assistance. Additionally, domain-specific datasets from partners such as Acxiom, Bloomberg, CoreLogic, Equifax, NIQ, and Transunion are now available on Google Cloud for AI training.

Secure Cloud

Google Cloud's permanent online AI partner, Duet AI, provides generative AI support for cloud users. It is integrated into three core solution areas - Mandiant Threat Intelligence, Chronicle Security Operations, and Security Command Center - to empower cloud users with generative AI capabilities. It also offers proactive search for hidden attacks based on frontline intelligence through Mandiant Hunt for Chronicle. These features are currently in preview and are planned for broader availability later this year. For more details, please refer to the draft blog post.

RECENT AI NEWS

RECENT AI TOOLS