Blogs

In the words of the famous Greek mathematician and philosopher Pythagoras, “Numbers Rule the Universe.” The figure below captures the news-worthy statistics of the recently concluded Google Cloud Next 2024, one of the biggest technology confabs in the world where technology experts like CXOs, decision-makers, technology leaders, practitioners, developers, and visionaries congregated to discover what’s next in technology—Cloud, AI, Data, and more—and its impact on the world.

1^st Time in Vegas! BTW it was announced during AWS re:Invent 2023

Extrapolating the magnitude of change and how the world could have transformed in an 8-month time span (from Google Next 2023), I write about my first-hand experiences at Google Next 2024 from Booth #S3, as part of Innova Solutions, the in the Security Area.

AI Everywhere

Unlike previous years at Google Next, when cloud infrastructure and broad technological advancements dominated the headlines, Google Next 2024 majorly focused on enterprise-grade AI technologies and their real-world applications, with Gemini taking center stage among the attendees. It is safe to assume that Gen AI was not only the backdrop for most of the event, but also the focal point of countless talks, product introductions, and demos. Not unexpected, as Google Cloud is presently used by ~90% Gen AI unicorns and above 60% funded Gen AI startups worldwide. This momentum reflects the increasing adoption of Gen AI across industries, demonstrating its transformative impact on businesses and applications, especially as you think about Gen AI taking over the productivity and experience agenda across large enterprises.

What Google Accomplished Between Google Next 2023 and 2024

It has been less than a year since Google’s last major cloud event, and Google utilized every second of the available time to achieve breakthrough feats that kept onlookers like us wondering if it was a turnaround in just 8 months from the previous event in 2023.

Recapitulating a few of those accomplishments:

• Introduced over 1,000 product advancements
• Google expanded its planet-scale infrastructure to 40 regions and invested in subsea cables to connect the world to Google’s Cloud with consistent low latency.
• Highlighted new, cutting-edge models, including Gemini models, and made them available to developers and enterprises.

Key Event Announcements

It would not have taken a billion-parameter large language model (LLM) to predict that Gen AI would be the major theme of Google Next 2024. Google rolled out new chips, software updates for AI workloads, updates to LLMs, and Gen AI-based assistants for its Machine Learning platform, Vertex AI.

At Google Next 2024, Google unveiled a series of key advancements across AI, security, and cloud infrastructure. Below is a consolidated list of the important announcements, in my view:

AI and Gen AI Enhancements

• The new version of Google’s open code generation model, Gemma, is now available on Vertex AI, Google’s AI platform to manage leading Gen AI models.
• Introduction of Code Assist, a tool for developers to write and analyze code more efficiently.
• Integrating Google’s Gemini AI assistant into fundamental cloud infrastructure solutions, such as security and databases.
• Launch of Google Vids, an AI-powered video creation app for work, as part of Google Workspace.
• Announcement of Imagen 2, extending Google’s image generation model to create short video clips from textual descriptions.
• Introduction of Axion CPUs for AI optimization.

Security Enhancements

• Gemini offers new services and capabilities for AI-driven security, such as Security Operations, Threat Intelligence, and Security Command Center.
• Introduction of new Gemini Cloud Assist IAM Recommendations, Key Insights, and Confidential Computing Insight features.
• Introduced Chrome Enterprise Premium, combining Chrome with Google’s threat and data protection, zero trust access controls, enterprise policy controls, and security insights/reporting.

Cloud Infrastructure Updates

• Improved AI Hypercomputer architecture with optimized hardware enhancements and storage for AI workloads
• Introduced Hyperdisk Storage Pools, a cost-effective block storage service ideal for AI inference.

Leading the Way Around Open-Source Models

Reinforcing its firm belief in the power of open technology to drive innovation for the greater good of all, Google encourages creativity and collaboration while maintaining developers’ access to technology and enabling customization as per use case. At Google Next 24, Google announced three open-source projects for developing and operating Gen AI models: MaxDiffusion, JetStream, and Optimum-TPU. Furthermore, the recent launch of Gemma, a new family of open-source AI models, demonstrates Google’s response to the growing demand for open-source AI models and acknowledges the competitive advantage of open-source AI, which is popular among programmers due to its customizability and cost efficiency.

With Google’s focus on building a state-of-the-art platform and AI ecosystem across the board, even non-Google models like Anthropic’s Claude 3 enjoy the same status as Google’s first-party models like Gemini Pro inside the Vertex Platform.

While an open-source foundation model can have many applications, code assistance and code generation are obvious choices for providing immediate value. This acted as a stepping stone for the launch of CodeGemma, a new open foundation model for code generation, code assistance, unit tests, and other code-related tasks. Gemma runs incredibly quickly on Apple’s latest M3 silicon, allowing developers to use it for free. Customers of Google’s Vertex platform will continue to have the freedom to choose the best model that meets their needs, with no lock-in.

Vertex AI Agent Builder

Google’s Vertex AI platform was a major focus at Google Next 2024, demonstrating how it helps businesses design and deploy Gen AI experiences. The platform includes tools like Agent Builder, an expansion and rebranding of Vertex AI Search and Conversation, Model Garden with 130+ models, including Llama, Gemma, and Model Builder, providing a comprehensive suite for developing and managing AI models.

To leverage the vast potential of Gen AI, enterprises need to overcome a sea of challenges, including balancing the rapid testing and iteration of AI models, apps, and agents with costs, governance, and performance at scale. To help customers overcome such challenges and accelerate the development of Gen AI agents, Google introduced Vertex AI Agent Builder, an expansion and rebranding of Vertex AI Search and Conversation solutions, as well as other upgraded developer tools.

NewsCorp is utilizing Vertex AI to search data from 30,000 global sources and 2.5 billion news stories updated daily. Other Google Cloud clients are building agents to improve internal efficiency. Researchers at the Mayo Clinic are leveraging Vertex AI Agent Builder to search for over 50 petabytes of clinical data.

Empowering Google with Gemini 1.5 Pro

Until Gemini 1.5 Pro, the largest context window in the world for a publicly available LLM was 200,000 tokens. Google has been able to extensively increase this, regularly processing up to 1 million tokens consistently, obtaining the longest context window of any large-scale foundation model. The window is available in Google AI Studio. Google AI Studio is the fastest way to build with Gemini models, and developers can easily integrate the Gemini API into their applications. It is available in 38 languages across 180+ countries and territories.

When evaluated on a wide range of text, code, image, audio, and video evaluations, Gemini 1.5 Pro outperforms Gemini 1.0 Pro on 87% of the benchmarks used in LLM development. In addition, with 1.5 Pro’s ability to process audio files, including audio-enabled videos, customers can process enormous amounts of data in a single stream, such as 1 hour of video (with subtitles in multiple languages), 11 hours of audio, codebases with over 30,000 lines of code, or over 700,000 words. This directly competes with OpenAI Sora.

Gemini 1.5 Pro powers new tools for coding faster, spotting cyber threats, processing audio files, and more.

Despite Gen AI still being in its initial stages, Google already has an interesting lineup of customers, such as Mercedes-Benz, IHG Hotels & Resorts, Best Buy, and Orange.

• Mercedes-Benz will collaborate with Google Cloud, using its AI and Gen AI technologies to advance customer-facing use cases across e-commerce, customer service, and marketing.
• IHG Hotels & Resorts is using Gemini to build a Gen AI-powered chatbot to help guests plan their next holiday via the IHG Hotels & Rewards mobile app.
• Best Buy is using Gemini to develop new and more convenient mechanisms for customers to fix product issues, reschedule deliveries, and others.
• Goldman Using Google Cloud and AI capabilities across multiple towers, from productivity to customer experience.

Enhance your Enterprise’s Cybersecurity Posture with AI

The quantity and gravity of cybersecurity threats continue to rise, and Gen AI can shift the scales in favor of defenders, with Security Agents assisting at all stages of the security lifecycle: Prevention, Detection, and Response. With new AI-driven innovations throughout Google’s security portfolio designed to deliver enhanced security outcomes, organizations are ready to onboard Google on their security team.

• Gemini in Threat Intelligence: Utilizes natural language to provide detailed insight into threat actor behavior. Gemini allows users to analyze far larger samples of potentially malicious code. Its larger context window enables analysis of module interactions, introducing fresh insights into the code’s underlying intent.
• Gemini in Security Operations: A new assisted investigation feature that converts natural language to detections, summarizes event data, proposes future actions to take, and guides users across the platform using conversational chat.
• Gemini in Security Command Center: Provides preview features that enable security teams to search for threats and other security events with natural language.

Leading global brands are already reaping the benefits of Google’s security strategies.

• Pfizer now analyzes data sources in seconds, compared to days in the past.
• 3Muses Gemini in Security Operations to reduce security noise.
• Fiserv’s Security Operations Center engineers can design detection and playbooks with less effort, leading to faster responses for analysts.

Google Cloud’s Investments & Expansions

Google Cloud is expanding its cloud regions to Columbus, Ohio, and Dallas, Texas, in the United States. Their expansion ambitions include South America, Europe, Asia, and the Middle East, ensuring faster and more efficient services for subscribers globally. Google is strongly committed to improving global connectivity, which includes a massive network of subsea cables that connect cloud infrastructure worldwide.

Vertex AI has expanded its grounding capabilities, including the ability to directly root responses using Google search. It also enhanced data residency assurances for data stored at-rest for Gemini, Imagen, and Embeddings APIs on Vertex AI to 11 new countries: Australia, Brazil, Finland, Hong Kong, India, Israel, Italy, Poland, Spain, Switzerland, and Taiwan.

Google also expanded its Cloud’s compute portfolio, with major product releases spanning compute and storage for general-purpose workloads, and for more specialized workloads such as SAP and high-performance databases. In addition, Google Cloud NetApp Volumes will expand to 15 new Google Cloud regions in Q2’24, with several enhancements.

What Was Missed Out: From the Perspective of an Onlooker

Google Next 2024 was quite unlike its competitors, and there were areas where I was hoping to explore more.

• Ethical Use of AI & Trust, since 2023, all the major tech conferences have begun and ended with emphasis on AI, and one of the single most important challenges is trust and ethics while using AI in the real world. Google Next, in my view, lacked enough conversation on the topic. In the absence of ethical guardrails, AI risks replicating real-world biases and discrimination, increasing divisions, and endangering fundamental human rights and freedoms. Similarly, the democratization of AI was not highlighted at the event except for open-source models, which demand less knowledge of AI from the user and may be used to build innovative AI software.
• A brief mention of sustainability was made at Google Next. Microsoft was the first one, and the others followed suit, but sustainability remains a huge challenge, and AI is clearly exacerbating the same with its need for power. However, AI is also capable of solving it and I hoped to see more of that at Google Next.

Let’s Not Forget, why was Innova there as a 1st time Sponsor

Over the last 18 months, we, at Innova Solutions, have gained substantial momentum in adopting Google Cloud and AI technologies, along with other hyperscalers and AI-focused technologies. The transition of Innova Solutions to becoming an AI services leader will eventually be a folktale at some point.

To begin with, we launched a few niche solutions for our customers looking to get more out of their Cloud, Google, and AI investments.

• Innova Knowledge BOT (until we can come up with a cooler name!), is our horizontally focused, knowledge base platform built with secure access to a corporate’s internal data sources for employees and corporate citizens to be self-sufficient. Our GenAI platform helps organizations empower their employees, reduce operations, and enhance the experience. The application areas for this knowledge platform are Legal, Finance, Sales, and HR. This platform is our showcase GenAI solution on Google Cloud with Google Vertex AI and other complimentary open-source technologies.
• Cyber Security Fusion Center with AI, our Security Compliance solution running on Google Cloud and Google AI, provides real-time monitoring, tracking, analysis, and threat intelligence for an enterprise in a rapid fashion with deep insights. Using AI, our solution can analyze historical data and provide better insights across all kinds of anomalies.
• Architecture Assessment and FinOps is a fully integrated end-to-end solution for analyzing your current cloud architecture, infrastructure investment, implementation best practices, and vulnerabilities. The solution’s tooling can help you navigate multi- and hybrid cloud settings without impeding your ability to develop a fully optimized cloud environment with improved application run time and room for new workloads at a considerably lower cost.