Google Cloud Next ’23 Highlights— AI and Beyond

Google Cloud Next is one of the most anticipated cloud computing events every year — and this year's event did not disappoint. Not surprisingly, the focus was on AI, machine learning, and generative AI. Google announced several new features and capabilities that will help businesses of all sizes take advantage of these cutting-edge technologies.

In this post, we cover the highlights of this year’s event. But first, before sharing a human’s view of the event, in keeping with this year’s AI theme, we thought we would run a little generative AI experiment just for fun. We asked three generative AI platforms to create a list of the event’s highlights for this blog post. Let’s see what they had to say — following by a human’s point of view. 

ChatGPT 3.5 test — lacked access to content

ChatGPT 3.5 did not deliver a lot of content. This is because the application does not have content dated prior to September 2021. Here’s its output:

Microsoft Bing Chat test — minimalist content generator

Microsoft Bing Chat performed better than ChatGPT 3.5 by providing a bit of relevant information. Here’s its output:

Google’s Bard test — the clear generative AI winner

Google’s Bard performed the best in this simple generative AI experiment. Here’s its output:

Now for the human-generated event highlights

Maybe the prompts just weren’t right, but we think we humans can do a bit better in sharing our favorite highlights from the event with you. Here goes.

Our first two picks focus on AI/ML — Vertex AI and Duet AI. These technologies are becoming increasingly important for businesses of all sizes and Google is ramping up the features to help organizations take advantage of them.

Vertex AI: This is Google’s unified platform for building, training, and deploying machine learning models.

These are the Vertex AI features we’re most excited about:

• Enterprise search on Generative AI App Builder and Conversational AI on Generative AI App Builder are now renamed Vertex AI Search and Conversation. They are generally available and can be used to build personalized and compelling generative applications with Vertex AI.
• Google is adding Meta’s Llama 2 and TII’s Falcon to Vertex AI, making it the first cloud provider to support first-party, open source and third-party models. APIs of these models will be accessible through the Model Garden on Vertex AI.

• Powered by Google DeepMind’s SynthID, Google announced Digital Watermarking on Vertex AI, a feature that embeds the watermark directly into the image of pixels, making it invisible to the human eye and difficult to tamper with.
• Not a direct feature, but Google Cloud has forged an interesting partnership with DocuSign to pilot a project to see how Vertex AI could be used to help generate smart contract assistants.
   

Duet AI: This AI collaborator application, launched during Google I/O, is now available in preview. It’s interesting and makes an appearance in several different places, starting with Google Workspaces for content generation, security analysis in Security Command Centre (SCC) and Mandiant, and integration into Big Query for a simplified query language.

Here are some of notable mentions of Duet AI:

• Duet AI is now part of Google Workspace and has a zero-cost trial. Integrated into Google Meet, it brings some interesting functions, like meeting summaries and recaps. People who don’t like attending too many meetings might rejoice as it allows a user to attend a meeting via a virtual bot that will take notes for them.
• Duet AI will help across the entire software development lifecycle, with the aim of minimizing context switching to help developers be more productive. In addition to code completion and code generation, it can assist with code refactoring and building APIs using simple natural language prompts. This will be a competitor to Microsoft® CoPilot
• Application and infrastructure operations, which often pour through documentation looking for configuration instructions, can chat with Duet AI in natural language across a number of services directly in the Google Cloud Console to retrieve how-to information about infrastructure configuration, deployment best practices, and expert recommendations on cost and performance optimization.
• Duet AI also makes an appearance in BigQuery. This is great for users who struggle with getting their SQL queries. It will provide contextual assistance for writing SQL queries, as well as Python code. It can generate full functions and code blocks, auto-suggest code completions and explain SQL statements in natural language. Also, it can generate recommendations based on schema and metadata. This sounds very cool and powerful!

Beyond AI

Along with updates to AI applications, Google also presented many more general announcements, including its Compute Engine, data and databases, and networking and security. Here are some that got our attention.

Compute Engine: We found some interesting updates in this area, including:

• General availability of A3 virtual machines from September 2023. A3 VMs will be powered by Nvidia H100 GPUs, and are purpose-built to train and serve demanding large-language models.
• The general-purpose Compute Engine family got a good refresh. It includes a third-generation Compute Engine instance, a next-generation Hyperdisk block storage, larger VMware Engine nodes and much more.
• A new GCE uptime SLA of 99.95% for memory-optimized VMs makes it the highest amongst hyperscalers.
• Future reservations, now in preview, is a new Compute Engine feature that allows the user to reserve compute capacity for up to one year in the future.
• GKE has a new enterprise version and a cool new multi-cluster feature called fleets that lets platform engineers easily group similar workloads into dedicated clusters, apply custom configurations per fleet, and even delegate cluster management to other teams.

Data and databases: You cannot leverage the power of AI and machine learning (AI/ML) without data. There weren't a lot of major new releases in this area, but, in our opinion, there were two big ones and one small one that are important. 

• AlloyDB Omni moves to public preview. By the way, you can read about a talk we did at Next ’23 on taking AlloyDB and Anthos on a multicloud ride using Omni.
• The fully managed Memorystore for Redis Cluster is available in preview. This is an open-source compatible service that provides up to 60 times more throughput than Memorystore for Redis, with microsecond latencies.

• If you’re invested in Cloud Bigtable and need a cross-region backup, you can now create a copy and store it in any project or region in which you have a Bigtable instance, and retain your backups for up to 90 days.

Networking and security: Security engineers might be a bit miffed that security is always mentioned last. We apologize but know we’re all used to it by now. Here are the most significant event mentions about security in our opinion:

• Powered by Palo Alto, and in preview, the Cloud Firewall Plus tier provides GCP-native firewalls with intrusion prevention capabilities. This is based on the current Cloud IDS product, which only provides intrusion detection.
• The TL;DR on Duet AI is that it has invaded the security landscape, including these features, which are all in preview:
  • Mandiant Threat Intelligence: It helps surface the most prevalent tactics, techniques and procedures (TTPs) used by threat actors against organizations by summarizing frontline threat intelligence.
  • Chronicle Security Operations: It aids in threat detection, investigation and response for cyber defenders by simplifying search, complex data analysis and threat detection engineering.
  • Security Command Center: It offers near-instant analysis of security findings and possible attack paths.
• Agentless vulnerability scanning, powered by Tenable, has been integrated into Security Command Center to detect operating system, software and network vulnerabilities on Compute Engine VMs.
• A win for private access, the internal Application Load Balancer now supports global access, which allows private clients from any Google Cloud region to access internal load balancers residing in any other Google Cloud region.
• Cloud Application Load Balancer now supports cross-project service referencing. If you’ve managed multi-project deployments where you are trying to use the Cloud Application Load Balancer from one Google Cloud project with a Cloud Function in another, you will relate to this new feature.

That wraps up our review of Google Cloud Next ’23. We hope you appreciated the human-generated version in our rapidly expanding AI world. In the future, the prompt-based AI tools will ultimately evolve into faster, simpler solutions, enabling modern organizations to solve a wide range of business problems and eliminating bottlenecks.

While Google Cloud has invested in making these technologies available to customers, generative AI needs both a strategic and tactical approach to the data platform, data lifecycle, ethics, security and more. Rackspace Technology® can assess your company’s workloads and recommend the best generative AI and Google Cloud solutions tailored to your specific needs.

We’ll see where the world of AI takes us next year at Google Cloud Next ’24 on April 9 to 11 in Las Vegas, NV.